New insights into chemical reactions with in situ 3D chemical mapping

There is a critical need, across many science and engineering disciplines (e.g., energy storage, microelectronics, as well as heterogeneous systems in environmental and life sciences), to correlate structural information (and its evolution) with the performance characteristics of materials. For example, lithium-ion batteries have been widely studied as major energy-storage devices. In these devices, electrochemically driven phase transformations—which directly affect the electrode performance—occur as the battery charges and discharges. It would therefore be highly desirable to track these chemical phase transformations as the cycling proceeds in an operating battery cell—in 3D and with adequate spatial and temporal resolution—i.e., to provide an in-depth understanding of the phase-transformation mechanism (which could lead to better design and optimization for electrode materials). It is challenging, however, to directly visualize and monitor the evolution of microstructural and chemical information (e.g., chemical composition distributions and chemical state changes) in 3D, under the operating conditions of the specimen.

In recent years, various characterization techniques have been used to study battery materials and to thus provide insight into the dynamic phase transformations of real battery electrodes as they cycle.^1–4 For instance, conventional synchrotron x-ray techniques (such as x-ray diffraction, scattering, and absorption spectroscopy) are powerful tools for in situ structural and phase-transformation studies. However, these methods are somewhat limited because the information obtained is often spatially averaged over the sample region (i.e., that is exposed to the x-rays). Scanning probes have better spatial resolution, but they can be quite time consuming (depending on the size of the sample being scanned). Furthermore, electron microscopy is a unique technique that provides high-resolution structural images, but the information along the depth axis is limited, and it is difficult to conduct under simulated real-battery working conditions. Synchrotron full-field, lens-based, hard x-ray, transmission x-ray microscopy (TXM) has therefore recently emerged as a powerful nondestructive tool that enables microstructural evolution to be monitored.^{5, 6} This technique has both high spatial-resolution (nanometer scale) and temporal resolution (milliseconds). In addition, with TXM it is possible to rotate the sample, which facilitates nanotomography measurements. In this way, detailed 3D structural information for complex systems can be obtained in in situ/operando conditions.

A newly developed TXM at the National Synchrotron Light Source (NSLS), at Brookhaven National Laboratory, has a number of unique features (including automated markerless tomography, local tomography, and constant magnification) that can be used to preserve the best-possible resolution.⁷ Furthermore, by combining synchrotron-based x-ray absorption near-edge structure (XANES) spectroscopy (which is highly sensitive to chemical and local electronic structures around selected elements) with TXM nanotomography, we have the ability to map chemical phase evolution at the nanometer scale (in 3D) at NSLS. Although there have been valuable studies of battery materials conducted in the past using a similar TXM-XANES technique,^8–11 the information was previously mapped in only two dimensions. To clarify ambiguities in the understanding of the phase-transformation mechanism, it is necessary to capture (in situ) the chemical evolution in 3D, but it is very difficult to acquire reliable images of a stable working battery cell over many days, and over 180^° rotations at different x-ray energies (i.e., that cross the absorption edge of the element of interest).

We have thus developed a full-field hard x-ray nanotomography spectroscopic imaging method. With this technique we can build five-dimensional data sets that track phase evolution in lithium iron phosphate (LiFePO₄) particles (i.e., the cathode material) while our in-house compact battery cell (which is compatible with the TXM setup and capable of repeated charging and discharging for many days) is charging. We are also able to scan the x-ray energy across the absorption edge of an element of interest.

In particular, in this work,¹² we scanned the x-ray energy (over a 100eV energy range) across the iron K-edge of LiFePO₄. At each energy point, we acquired hundreds of images (2D projections) over a full 180^° rotation of the sample. We then reconstructed these 2D projections into a 3D structure, and thus formed a series of 3D reconstructions at different x-ray energies. At the same position in the series of 3D reconstructions, each voxel corresponds to a XANES spectrum (i.e., if the detector has 2048 × 2048 pixels, the 3D reconstruction would have 2048 × 2048 × 2048 voxels, or more than 8 billion XANES spectra). Finally, we fit each XANES spectrum to obtain the chemical phase information for each voxel position and thus produced a complete 3D chemical phase map.

We also repeated this procedure at different charging stages to reveal the chemical phase transformation as a function of the charging time. We illustrate how the chemical phase transfers from LiFePO₄ to FePO₄, inside a LiFePO₄ particle, as the battery cell is charging in Figure 1. These images show that the chemical phase transformation is anisotropic at the beginning of charging, and that it changes to isotropic during the highly charged state. The method that we have developed has thus led to the direct observation of how the phase transformation (in 3D) occurs, and shows whether a new/intermediate phase forms during the phase-transformation process (i.e., providing key and precise insights into the processes that occur inside a battery electrode). Our results also help to clarify previous ambiguities regarding the mechanism of phase transformation in a LiFePO₄ electrode, and they point to potential ways to improve the performance of the cell.

Figure 1. Illustration of how the chemical phase inside a lithium iron phosphate (LiFePO₄) particle evolves with increasing charging time. The cut views of the particles reveal the change from anisotropic to isotropic phase-boundary motion. Scale bar indicates 10μm.

In summary, we have developed a new in situ, 3D, full-field, hard x-ray nanotomography technique to track the phase evolution of lithium-ion batteries. Such in situ, 3D chemical mapping of a system, under operating conditions, provides new insights into how microstructures in a battery electrode can affect chemical phase transformation. Our work also provides a better understanding of the chemical reaction pathways and how they are correlated to battery performance, and will thus lead to better designs of advanced battery materials. Beyond battery research, our methodology is applicable to many fields of science and engineering (e.g., in energy, chemistry, physics, and materials sciences). To gain further understanding of the origin of the phase transformation, as well as its propagation/distribution, we will next work on modeling and simulations to investigate how the strain forms and how lithium-ion diffusion occurs during electrochemical reaction.

This work was performed by the author's team, including Jiajun Wang, Yu-chen Karen Chen-Wiegart, and Christopher Eng. The work also benefited from in-depth discussions with Qun Shen. Use of the NSLS was supported by the US Department of Energy, Office of Basic Energy Science (contract DE-AC02-98CH10886).