Berkeley Lab Science Beat Magazine banner Berkeley Lab logo
February 17, 2004
Testing Better Batteries

Building a better battery is a key goal for those who would like to see electric and hybrid-electric vehicles become viable options in the car market. However, progress toward this goal has been slow. Many labs are seeking battery anode (negative electrode) and cathode (positive electrode) materials that will last longer, suffer less degradation, and operate safely over wider temperature ranges than is currently possible.

Hybrid-electric vehicles combine the internal combustion engine of a conventional vehicle with the electric motor and batteries of an electric vehicle, achieving twice the fuel economy of conventional vehicles.

As part of this battery research effort, a unique cell development program has been under way in Berkeley Lab's Environmental Energy Technologies Division (EETD), led by Kathryn Striebel. The project uses standardized cells to assess, in a working battery, the performance of promising new materials. The project aims to bridge the gap between materials research and commercial battery development.

EETD has long studied advanced materials for batteries. The work is currently funded by the Department of Energy's Batteries for Advanced Transportation Technologies program (BATT), of the Office of FreedomCAR and Vehicle Technologies. BATT, administered for DOE by EETD, consists of six research tasks involving Berkeley Lab and a number of other institutions and national laboratories.

Standard cells test realistic conditions

"The idea of this research element is to take new materials from labs and build them into test cells for new batteries," says Striebel. "We build new materials from different sources into these test cells and run a set of standard tests to see how they perform under realistic conditions. Then we disassemble the test cells and, after some additional electrochemical testing of our own, we send samples to the Berkeley Lab researchers focusing on diagnostic techniques, such as Raman spectroscopy, Fourier-transform infrared spectroscopy, and many others."

Experimental materials come from labs all over the world, including EETD's own electrochemistry labs. The testing helps determine why electrode materials fail or degrade. To be successful, a battery for automotive applications must meet DOE criteria for features such as weight, cost, power density, and operating temperature range.

These criteria include a 10-year life, a cost of $150 per kilowatt-hour or less, the ability to operate between minus-40 and plus-50 degrees Celsius, and a lifetime loss of capacity of no more than 20 percent. Batteries for hybrid-electric vehicles differ slightly from those for electric vehicles in that they also need to be able to provide numerous pulses of power for acceleration, as well as accept charge during regenerative braking.

Currently, lithium-ion-based cells show promise for meeting these performance goals. One candidate chemistry for lithium-ion batteries is based on lithium iron phosphate (LiFePO4) and natural graphite (NG).

"The central goal for us," Striebel says, "is to determine which materials work the best — and when they fail, to answer the question 'why?'"

Lithium iron phosphate material has some advantages: it is stable and flame retardant, it has a long cycle life, and it shows promise for meeting the goal of no more than 20- percent capacity degradation over the battery's lifetime. However, the capacity of LiFePO4 batteries is currently insufficient for use in vehicles.

Indeed, no material currently meets all of DOE's goals for automotive batteries. One important reason is that the performance of existing materials degrades significantly after many charge-discharge cycles. "Our strength is in our understanding of degradation mechanisms in battery materials," Striebel says of EETD. "If we can nail down the mechanisms of degradation, it will be a great help to everyone working in the field."

Test pouches

The program's test cells are small, thin pouches just 12 square centimeters in area and hardly more than an inch on a side (about 3.5 centimeters), which can store an average of 12 milliampere-hours of charge.

The effort to make a cell starts with 5 to 20 grams of an experimental material—an amount considered large for a new material, which may exist only in tiny quantities in a single lab. The material is mixed with carbon, a binder polymer, and a solvent to form a slurry. This slurry is cast in thin layers onto a foil current collector, then dried extensively.

A few grams of experimental battery material are mixed with carbon and cast on a foil current collector (left) to make test-cell pouches 3.5 centimeters on a side (center). Up to 64 cells are tested simultaneously (right), under conditions like those in a hybrid-electric vehicle battery.

One anode and one cathode are placed in a flexible pouch with a porous separator and transferred to a helium-filled glove box for finishing. At this point, electrolyte is added, and the pouch is sealed to protect the cell from water vapor during testing. The pouch is then compressed and mounted on a test stand, usually along with many other cells undergoing testing.

The tester can charge and discharge up to 64 test cells simultaneously, according to any specification. For example, it can run through continuous charge-discharge cycling at constant current, letting the cells rest between half-cycles—which is the procedure for determining baseline cell performance—or it can charge and discharge with short, high-current pulses, simulating the conditions that a hybrid-electric vehicle's battery might encounter.

The tester measures current, voltage, and other parameters, and for each test cell provides impedance characteristics, capacity, and power as a function of time or number of cycles. After a cell reaches a predetermined end-of-life limit (low capacity or power), additional diagnostic cycles are carried out before the cell is removed to the glove box for disassembly.

Once the cell is disassembled, Striebel and her colleagues may subject the experimental material to a range of additional tests to investigate its degradation mechanisms. These tests might use Raman spectroscopy, Fourier-transform infrared spectroscopy, and other spectroscopic methods; X-ray diffraction; and transmission electron microscopy.

"The testing is an ongoing program," says Striebel. "We continue to test new materials as they are developed. The results allow us to compare the performance of different materials with one another." Striebel's group has also been working with that of John Newman of EETD and the University of California at Berkeley, developing computer models of battery performance. "This really helps us isolate why these materials perform the way they do."

Test results are presented at U.S. and international meetings and published in peer-reviewed journals, so the data are available to the scientific community as well as to battery developers. "Recently, we used the computer modeling directly to help in the comparison of six different sources of LiFePO4 from around the world," says Striebel. "This approach generated a lot of interest at a recent meeting of the Electrochemical Society."

Additional information