Unraveling structural ambiguities in lithium-rich transition metal oxides

October 29, 2015
Source:

Using complementary microscopy and spectroscopy techniques, researchers at Lawrence Berkeley National Laboratory (Berkeley Lab), Calif., say they have solved the structure of lithium- and manganese-rich transition metal oxides, a potentially game-changing battery material and the subject of intense debate in the decade since it was discovered.

Researchers have been divided into three schools of thought on the material's structure, but a team led by Alpesh Khushalchand Shukla and Colin Ophus spent nearly four years analyzing the material and concluded that the least popular theory is in fact the correct one. This material is important because the battery capacity can potentially be doubled compared to the most commonly used Li-ion batteries today due to the extra lithium in the structure. "However, it doesn't come without problems, such as voltage fade, capacity fade, and DC resistance rise," says Shukla. "It is immensely important that we clearly understand the bulk and surface structure of the pristine material. We can't solve the problem unless we know the problem."

A viable battery with a marked increase in storage capacity would not only shake up the cell phone and laptop markets, it would also transform the market for electric vehicles (EVs). "The problem with the current lithium-ion batteries found in laptops and EVs now is that they have been pushed almost as far as they can go," says Ophus. "If we're going to ever double capacity, we need new chemistries."

Using state-of-the-art electron microscopy techniques at the National Center for Electron Microscopy (NCEM) at Berkeley Lab's Molecular Foundry and at SuperSTEM in the UK, researchers imaged the material at atomic resolution. Because previous studies have been ambiguous about the structure, the researchers minimized ambiguity by looking at the material from different directions, or zone axes. "Misinterpretations from electron microscopy data are possible because individual two-dimensional projections do not give you the three-dimensional information needed to solve a structure," Shukla says. "So you need to look at the sample in as many directions as you can."

Scientists have been divided on whether the material structure is single trigonal phase, double phase, or defected single monoclinic phase. The "phase" of a material refers to the arrangement of the atoms with respect to each other; Ophus, a Project Scientist at the Molecular Foundry, explains how easy it is for researchers to reach different conclusions: "The two-phase and one-phase model are very closely related. It's not like comparing an apple to an orange—it's more like comparing an orange and a grapefruit from very far away. It's hard to tell the difference between the two."

In addition to viewing the material at atomic resolution along multiple zone axes, the researchers made another important decision, that is, to view entire particles rather than just a subsection. "Imaging with very high fields of view was also critical in solving the structure," Shukla said. "If you just look at one small part you can't say that the whole particle has that structure."

Putting the evidence together, Shukla and Ophus are fairly convinced that the material is indeed defected single phase. "Our research gives strong support for the defected single-phase monoclinic model and rules out the two-phase model, at least in the range of compositions used in our study," says Ophus, whose expertise is in understanding structure using a combination of computational methods and experimental results.

"We need to know what goes on at the atomic scale in order to understand the macroscopic behavior of new emerging materials, and the advanced electron microscopes available at national facilities such as SuperSTEM or NCEM are essential in making sure their potential is fully realized," he adds.

In addition to solving the structure of the bulk material, which has been studied by other research groups, they also solved the surface structure, which is different from the bulk and consists of just a few layers of atoms on select crystallographic facets. "The intercalation of lithium starts at the surface, so understanding the surface of the pristine material is very important," Shukla says.

On top of the STEM (scanning transmission electron microscopy) imaging that they used for the bulk, they had to use additional techniques to solve the surface, including EELS (electron energy loss spectroscopy) and XEDS (x-ray energy dispersive spectroscopy). "We show for the first time which surface structure occurs, how thick it is, how it's oriented in relation to the bulk, and in particular on what facets the surface phase does and doesn't exist," Ophus says.

An important part of the study was the quantity and quality of the samples studied. They started with lab-made samples. They used a molten-salt method that produces high-quality discrete primary particles that are impurity-free, making them ideal candidates for performing fundamental characterization. Taking a conservative approach, the researchers also decided to procure and analyze two commercial samples from two different companies.

"We could have finished the paper a year earlier, but because there was so much controversy we wanted to make sure we didn't leave any stone unturned," says Shukla. In the end, it took nearly four years to complete the research. Ophus calls it a "tour de force of microscopy" because of its thoroughness.

The work was funded by the Vehicle Technologies Office under the U.S. Department of Energy. The Molecular Foundry is a DOE Office of Science User Facility.

Image caption — The cube on the right represents the structure of lithium- and manganese- rich transition metal oxides. The models on the left show the structure from three different directions, which correspond to the STEM images of the cube.