Developing the science and technology for next-generation battery systems has long been a focus of research at Lawrence Berkeley National Laboratory, dating back to the early 1980s. Lithium-ion batteries (sometimes abbreviated Li-ion) are the primary focus of current research, because their light weight and high energy-density make them ideal candidates for transportation use.

High-performance rechargeable batteries will be essential to the success of all kinds of electric vehicles.

The Department of Energy's Office of FreedomCar and Vehicle Technologies is supporting researchers in the Lab's Environmental Energy Technologies Division (EETD) who are developing high-performance rechargeable batteries for use in a veritable alphabet soup of transportation: electric vehicles (EVs), hybrid-electric vehicles (HEVs), plug-in hybrid-electric vehicles (PHEVs), and fuel-cell electric vehicles (FCEVs).

Batteries and other energy storage technologies are critical to advanced, fuel-efficient transportation — so much so that they are one of DOE's Energy Strategic Goals. The automotive industry is working together with the FreedomCAR and Vehicle Technologies Program to identify technical barriers to improving energy storage technologies; DOE-funded research is aimed at toppling these barriers.

The BATT Program, Batteries for Advanced Transportation Technologies, is a $6 million program being carried out at Berkeley Lab and other institutions to research fundamental problems, those chemical and mechanical instabilities which have impeded the development of EV, HEV, PHEV, and FCEV batteries with acceptable costs, performance, lifetimes, and safety. The aim is to better understand battery cell performance and the factors that limit battery lifetime.

Microscale and nanoscale probes

The third lightest element on the periodic table, following hydrogen and helium, is lithium — a rising star in battery chemistry. Lithium-ion batteries are considered the state of the art, the future of battery technology. Energy is stored in these batteries through the movement of lithium ions between the cathode, or positive terminal, and the anode, or negative terminal, electrodes which effectively "house" the ions. (Ions are charged particles, in this case atoms with net positive charge.)

For transportation purposes, lithium's very light weight can provide a substantial savings compared to batteries made of heavier metals. Another big advantage of Li-ion chemistry is that compared to aqueous batteries such as lead acid, nickel metal hydride, or nickel cadmium, it yields high open-circuit voltage — the higher the voltage, the higher the power and the better the acceleration.

	One technology with great promise for future electric vehicles is the light-weight, high-voltage lithium-ion battery widely used in today's electronic devices.

Lithium ion batteries are already among the most popular for portable electronics, having a superior energy-to-weight ratio and a slow loss of charge when not in use. To be commercially viable for transportation, however, Li-ion batteries will need to last 10 to 15 years; their cost will have to be significantly reduced and their safety improved.

"Unfortunately, Li-ion batteries are thermodynamically unstable," says EETD researcher Robert Kostecki. "Sooner or later, they are destined to fail."

Basic physical law in effect determines that you can't keep recharging a battery indefinitely without gradual loss of charge capacity and power capability. The active materials in the electrodes tend to react chemically with the electrolyte. The reaction products form protective surface films on electrode surfaces, which stop or slow down these surface reactions.

Over the long term, during charge/discharge cycling and storage, the instability of these surface films, the intrinsic behavior of battery components, and variations in engineering quality contribute to both power fade and capacity loss in conventional lithium-ion-cell chemistries.  

"We need to detect and describe these processes to understand their mechanism and kinetics before we can find ways to slow them down," says Kostecki. "Interfacial phenomena often manifest themselves at nano- or microscales and can be detected and characterized only by surface-sensitive techniques."

Materials diagnostics is thus a major part of the effort to develop the next generation of Li-ion batteries for transportation. Kostecki and his colleagues are among first to use a set of unique microscale and nanoscale techniques to study the degradation processes, and their work is leading to better materials and manufacturing techniques for advanced batteries.

Correlations between microscopic phenomena and macroscopic behavior

"Researchers routinely use a number of standard techniques to study battery chemistry," Kostecki explains. Electrochemical techniques can monitor changes of battery charge/discharge characteristics, for example, and x-ray spectroscopy can detect chemical and structural changes in an electrode's active materials.

Current-sensing atomic force microscopy is used to measure both the conductance and the surface topography of electrode materials, with resolution at micrometer and nanometer scales.

But these methods usually cannot sense local phenomena in the electrodes, which take place at the microscale (measured in millionths of a meter) or even the nanoscale (billionths of a meter). Kostecki and his Berkeley Lab colleague Frank McLarnon were among the first to apply instrumental methods allowing them to monitor the composition and structural changes of battery electrodes at nanoscale or microscale resolution.

Using current-sensing atomic force microscopy (CSAFM), Kostecki and McLarnon studied the surface and electric conductance of composite electrodes used in various lithium-ion batteries. A single scan of the conductive AFM tip over the cathode surface produces two images simultaneously, a topographic image and a conductance image.

The tip of the current-sensing atomic force microscope is in physical contact with the oxide. The magnitude of the current is determined by the local electronic properties of the electrode and the tip, the voltage difference between tip and sample, and the geometries of the CSAFM tip and the local electrode surface.

The researchers also used Raman microscopy to carry out a microanalysis of the electrode surface. Raman microscopy, a spectroscopic technique used in physics and chemistry, measures laser light scattered from the sample to provide information about its chemical composition and structure. By collecting thousands of Raman spectra from small sections of electrode surfaces, they were able to produce and compare unique color-coded surface composition maps of electrodes from Li-ion cells.

	Microscopy using Raman spectroscopy techniques can be used to create maps that yield high-resolution information about the chemical composition and structure of electrode materials.

"Our diagnostic evaluations of composite electrodes revealed changes in electrode surface composition, structure, electronic conductivity, and local state of charge, which accompany cell cycling and aging," says Kostecki. "Our hypothesis is that the phenomena that cause degradation in batteries occur at micro- and nanometer scale and can only be detected with appropriate microscopic techniques. To detect them, we have developed and applied techniques and methodologies never used before in this field. We were the first to use high-resolution Raman microscopy mapping, which revealed the nonuniform distribution of the electrode state of charge at a micrometer scale. The data allowed us to identify the local processes that contribute to significant loss of electronic conductivity within the electrode and consequently to the capacity loss."

A dopant effect is not what it seems

Kostecki applied these local-probe techniques to investigate lithium iron phosphate, LiFePO4, considered one of the most promising cathode materials for the next generation of Li-ion batteries.

However, says Kostecki, "The poor electronic conductivity of LiFePO4 compared to transition-metal oxide cathodes is a serious limitation on its use in high-power Li-ion systems. Controversial reports in the literature suggested different pathways to improving its electrochemical performance. The lack of understanding of the LiFePO4 operating mechanism has delayed introducing this material into a new generation of Li-ion batteries."

Kostecki and his colleagues performed CSAFM and Raman microscopy on two samples of lithium iron phosphate powder, one pure and one that had been one-percent doped with niobium by a team of MIT scientists, replacing some lithium atoms with niobium atoms in an attempt to improve the compound's electronic conductivity.

In the CSAFM images, the researchers detected no electronic conductance in pristine lithium iron phosphate at any location; the conductance image was pure white. The niobium-doped sample, on the other hand, had better electronic conductivity, at first suggesting that the niobium was indeed responsible for adding conductivity.

A closer look at the niobium-doped sample raised questions, however. In the conductance image, black splotches revealed numerous small sites with good conductance, scattered across the surface — but curiously, on the corresponding morphology image, the researchers identified grains of active electrode material that had become completely insulating. Indeed, conductivity was very nonuniform, localized mainly in deep crevices and pockets between agglomerates.

Typical Raman spectra of lithium iron phosphate powder showed that the material consisted not just of lithium iron phosphate but of iron oxides, phosphides, and elemental carbon impurities as well. The Raman microscopy images of the same powders, niobium-doped and pure, revealed that the carbon content in the niobium-doped sample was much higher than in the pristine lithium iron phosphate.

Panels at left are CSAFM measurements of the surface topography and conductance of lithium iron phosphate, a sample doped with niobium shown at top and a sample of the pure state at bottom. The doped sample has a different topography, and it conducts electricity better but unevenly (black regions indicate high conductance). In the center panel, a Raman spectrum of a sample reveals iron oxides, phosphides, and elemental carbon impurities in addition to the lithium iron phosphate compound. Raman microscopy images, right, reveal that the niobium-doped sample at top, which is the better conductor, contains much more carbon (blue areas) than the pristine lithium iron phosphate sample at bottom.

An organic, niobium-containing precursor used in the doping process was the source of the extra carbon. The researchers observed that the carbon distribution in the niobium-doped sample corresponded exactly to the pattern of conductivity observed in the CSAFM images. They concluded that it was actually the carbon additive, not the niobium, which was responsible for the doped material's conductivity increase and better electrochemical performance.

"We would not have been able to reach this conclusion without the unique combination of nanoprobe techniques and innovative methodologies that we applied to study this system," says Kostecki. "We determined that the key to increasing the conductivity of the material and making it a more effective cathode material was to incorporate more conductive carbon — and to improve the distribution of carbon deposits. This shifted the focus away from materials science toward better engineering."

Improving battery longevity

Kostecki applied the same set of instrumental methods to study the mechanism of Li-battery degradation over time. First, using Raman and CSAFM imaging, he and his colleagues characterized the surface chemical composition, structure, morphology, and electronic conductivity of a fresh Li-ion electrode. The imaging allowed them to identify the original distribution of the electrode components — the active material and carbon additives at the surface of the electrode. They reexamined the electrode with the same tools after prolonged charge/discharge cycling and storage, looking for changes that might be linked to detrimental surface phenomena responsible for the loss of electrochemical performance.

The Raman images showed a marked change in the material's structure as well as its surface composition and distribution. While some areas of the sample remained relatively unchanged, elsewhere there were large changes in both surface structure and composition — the more active material was exposed and less carbon additive was present.

"Loss of surface electronic conductivity accompanied the observed changes in the surface chemistry," Kostecki explained. As a result, some particles of the electrode active material became partially or fully electronically disconnected from the rest of the electrode and become inactive. "These highly localized phenomena had severe impacts on the overall electrochemical performance of the electrode and the whole Li-ion battery. Its charge capacity was diminished and impedance significantly increased."

Kostecki says, "The nano- and microprobe analytical tools allowed us to demonstrate that the localized deactivation processes that occur on a microscopic scale can be directly linked with the macroscopic behavior. It was the first time these techniques were applied in such an efficient and concerted way to study battery surface phenomena. The results of these studies have given us a better understanding of both the nature of the process, and some ideas about how to prevent them or slow them down."

It has also motivated materials scientists and battery engineers to work together more closely, so that the dream of cheaper, longer lasting, and safer lithium batteries for advanced electric vehicles becomes a reality sooner.

Additional information

• More about the Berkeley Electrochemical Research Council
• More about BATT, Batteries for Advanced Transportation Technologies