In situ characterization—Material synthesis

Material chemistry developments play a vital role in the advancement of Li-ion batteries. High-temperature solid-state synthesis is still widely used for making lithiated cathode materials, because the processing can be easily scaled up for industrial manufacturing. Depending on the synthesis routes used, cathode materials with the same nominal composition can have distinctly different electrochemical performance. Therefore, an in-depth understanding of the nucleation, crystallization, and growth processes associated with synthesis is required to discern and manipulate the properties of cathode materials.^{Reference Jensen, Bremholm, Nielsen, Joensen, Pedersen, Birkedal, Chen, Almer, Søgaard, Iversen and Iversen6} HEXs are particularly well-suited for in situ monitoring of solid-state reactions and phase transformations because they can penetrate through complex sample environments, such as steel containment vessels for high-pressure applications and specially designed high-temperature furnaces.

For example, high-energy small- and wide-angle x-ray scattering (HE-SAXS/WAXS)^{Reference Jensen, Bremholm, Nielsen, Joensen, Pedersen, Birkedal, Chen, Almer, Søgaard, Iversen and Iversen6} has recently been used to reveal TiO₂ nanoparticle nucleation and growth behavior under synthesis-reaction conditions, where TiO₂ is a precursor anode material. A schematic of the setup used is shown in Figure 2a. In a similar manner, HEX diffraction has been used to characterize the formation of olivine LiFePO₄in situ,^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7} a nickel-rich full-concentration gradient oxide Li(Ni_0.6Mn_0.2Co_0.2)O₂,^{Reference Li, Xu, Ren, Lu, Wu, Wang, Miller, Sun, Amine and Chen8} the migration mechanism of transition metal ions in spinel Li(Ni_0.5Mn_1.5)O₄,^{Reference Xu, Qin, Ren, Cai, An, Amine and Chen9} and in situ monitoring and control of graphene-modified nanostructured vanadium pentoxide V₂O₅ hybrids with extraordinary electrochemical performance for Li-ion batteries.^{Reference Liu, Li, Liu, Zhang, Ren, Sun, Lu, Zhou, Stanciu, Stach and Xie10}

Figure 2.

(a) Experimental setup for real-time, in situ high-energy small-angle x-ray scattering/wide-angle x-ray scattering (SAXS/WAXS) studies of supercritical reactions in nanomaterials used for batteries. Incident x-rays pass through the reactor window during the course of the experiment, with the x-ray scattering from nanoparticles present in the reactor collected with the WAXS detector, while scattering from the fiber material is collected with the SAXS detector, which is placed at a shallower angle relative to the transmitted beam than the WAXS detector. The temperature and amount of the reactants in the reactor can be varied while x-rays pass through the reactor, allowing for real-time in situ observations. (b) Development of crystallinity can be monitored by WAXS, as exemplified by the anatase 200 reflection. The SAXS data (c) provide information on mesostructure and (d) enable the extraction of the particle-size distribution (diameter) using a hard-sphere model. The colors shown in (b–d) correspond directly to the vertical axes in each respective plot, where dark blue represents the lowest values and dark red the highest. Reprinted with permission from Reference Reference Jensen, Bremholm, Nielsen, Joensen, Pedersen, Birkedal, Chen, Almer, Søgaard, Iversen and Iversen6. © 2007 Wiley. Note: Q, magnitude of the scattering vector of scattered x-rays in reciprocal space, in Å^–1.

In situ characterization—Thermal stability

With increasing demand for more powerful and energy-dense battery materials, understanding and reducing thermal instability in electrode and electrolyte materials is essential for maintaining and improving safety and failure tolerance in existing and emerging batteries. Overcharging, overdischarging, electrical short circuits, and external effects can all lead to a rapid rise in temperature that may induce thermal runaway and ignite flammable materials such as the electrolyte. Additionally, many cathode materials are unstable or exhibit poor reversibility and lifetimes during cycling at elevated temperatures, especially during high-power (dis)charging. Determining these thermal instability mechanisms, and the interactions between these materials and electrolytes, is therefore essential toward not only creating safe batteries, but also increasing performance and lifetimes.

As discussed with in situ synthesis studies, HEXs are well suited to characterize battery materials contained within sample environments due to their high penetrating power. For example, in situ HEX diffraction experiments have been combined with differential scanning calorimetry to explore structural changes and thermal degradation of novel full-concentration gradient (FCG) transition-metal oxide cathodes, Li(NiMnCo)O₂, where the respective concentration of Co varies from a maximum in the center of a particle to near zero at the outer surface, and the opposite concentration gradient exists for Mn.^{Reference Sun, Chen, Noh, Lee, Jung, Ren, Wang, Yoon, Myung and Amine11} Sun et al. combined diffraction and calorimetry to deduce how the FCG structure improved both thermal stability and overall performance and lifetime over traditional Li(NiMnCo)O₂ cathodes.^{Reference Sun, Chen, Noh, Lee, Jung, Ren, Wang, Yoon, Myung and Amine11} In FCG cathodes, a manganese-rich surface added thermal stability to a nickel-rich core without significantly lowering capacity, when compared with the pure nickel-rich core composition. Additionally, the FCG cathodes’ intraparticle nanorod morphology increased lithium diffusivity and thus overall rate capability at a 5C rate, where 1C corresponds to the current required to fully charge an electrode in 1 h, and 5C corresponds to the current required to fully charge an electrode in 1/5 h, or 12 min.^{Reference Sun, Chen, Noh, Lee, Jung, Ren, Wang, Yoon, Myung and Amine11} A schematic of the sample environment for the experimental setup, as well as changes in cathode structure with respect to temperature, are shown in Figure 3.

Figure 3.

(a) Photograph of a standard stainless-steel high-pressure crucible for in situ high-energy x-ray (HEX) scattering diffraction study of electrode and electrolyte reactions during thermal ramping. The flower-shaped inset part with a 1-mm diameter tunnel is made of either Cu for holding anode materials or Al for cathode materials.^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7} The inset is a schematic showing how HEXs penetrate through the stainless-steel cap to probe the sample inside the cell and scatter out from the other side to be collected by a 2D detector. Contour plots of the in situ HEX diffraction patterns in a characteristic region of (b) delithiated core–shell materials, and (c) delithiated full-concentration gradient materials during thermal ramping from room temperature up to 375°C. In (b), an abrupt phase transition occurs at about 210°C, suggesting poor thermal stability, while in (c), this transition occurs over a much wider temperature range, suggesting significantly improved thermal stability. The color contours correspond to the scattered x-ray intensity for a given scattering vector magnitude, where blue corresponds to low intensity and red to high intensity with other colors falling in between on a linear scale. Reprinted with permission from Reference Reference Sun, Chen, Noh, Lee, Jung, Ren, Wang, Yoon, Myung and Amine11. © 2012 Macmillan Publishers Ltd. Note: T, crucible temperature during in situ characterization; Q, magnitude of the scattering vector of scattered x-rays in reciprocal space, in Å^–1.

Chen et al. similarly identified several thermal decomposition pathways for various chemical environments in delithiated Li_1–x[Ni_1/3Mn_1/3Co_1/3]_0.9O₂.^{Reference Chen, Ren, Lee, Johnson, Qin and Amine12} More generally, the combination of diffraction and heating in situ has been used to evaluate the stability and thermal abuse tolerance of a novel nonaqueous Li₂B₁₂F_12–xH_x electrolyte,^{Reference Chen, Ren, Jansen, Lin, Weng and Amine13} and to study the structural changes and thermal stability of electrodes without the presence of electrolytes.^{Reference Yang, Sun and McBreen14}

In operando characterization—Time-resolved studies

The penetration power of HEXs allows for in operando characterization on a variety of battery formats and with high-spatiotemporal resolution. During the electrochemical cycling of a battery material, various phase transitions occur between different states of charge, with some of these transitions becoming metastable or irreversible, or altering phase-change pathways as the rate of (dis)charge changes. Conventional ex situ techniques are unable to examine metastable phases or effectively determine rate-sensitive changes in transition pathways, and often have great difficulty in describing the formation of irreversible phases. Thus, in operando experiments are particularly powerful for elucidating the evolution of many battery materials during electrochemical cycling and understanding the nature of phase changes associated with (dis)charge processes, especially when these changes are transient, metastable, or rate-sensitive in nature.^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7}

In operando HEX experiments have revealed the mechanisms behind rate dependence and the limitations of phase-change processes in cathode materials. Yoon et al.^{Reference Yoon, Nam, Jang, Chung, Cho, Choi, Hanson and Yang15} observed for a LiMn₂O₄-LiNi_1/3Co_1/3Mn_1/3O₂ composite cathode that LiMn₂O₄ was able to maintain useable capacity at fast cycling rates (4C, or four full charges in 1 h) by employing an additional phase-transformation route with the formation of a second distinct phase that augments the pseudo-one-phase transition behavior also observed at slower rates (0.5C, or one full charge in 2 h). In contrast, LiNi_1/3Co_1/3Mn_1/3O₂ was unable to maintain capacity at faster rates using only one phase-transition mechanism. The phase-change behavior of LiFePO₄-based cathodes during cycling has also been studied in more detail thanks to the capabilities of HEX,^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7} including examining commercially produced LiFePO₄-based batteries.^{Reference Liu, He, Li, Liu, Ren, Lu, Lu, Stach and Xie16} While lively debate as to the nature of (de)lithiation in LiFePO₄ continues in the literature, in operando studies have revealed rate-sensitive transitions and, in certain cases, metastable phases that were previously unable to be studied. H. Liu et al.^{Reference Liu, Strobridge, Borkiewicz, Wiaderek, Chapman, Chupas and Grey17} and Q. Liu et al.^{Reference Liu, He, Li, Liu, Ren, Lu, Lu, Stach and Xie16} observed a nonequilibrium solid-solution reaction mechanism in LiFePO₄ under high cycling rates, which may explain the excellent rate capability of LiFePO₄-based cathodes.

In addition to examining cathode material issues, HEXs are useful for studying new anode materials. Many of the most promising anode materials to replace graphite employ materials that undergo sharp phase transitions and dramatic volume changes during (de)lithiation.^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7} HEXs have enabled the close study of the amorphization of these anodes during the first cycle,^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7} as well as quantification of lithiation-induced strains under galvanostatic conditions.^{Reference Glazer, Cho, Almer, Okasinski, Braun and Dunand18} Wang et al.^{Reference Wang, Wu, Key, Yang, Grey, Zhu and Graetz19} demonstrated that nanoscale Si undergoes a two-step amorphization process, as shown in Figure 4, which is in contrast to the linear amorphization that occurs in micrometer-sized Si particles.

Figure 4.

(a) In operando x-ray diffraction patterns for crystalline silicon taken at various different capacities during the first lithiation cycle with (b) corresponding electrochemical curve, and (c) proposed mechanism for the first cycle amorphization of a crystalline silicon nanoparticle (red) reacting with lithium (blue). (a) As lithiation occurs (from bottom to top), the heights of both Si diffraction peaks decrease while the widths increase, indicating an increase in disorder and amorphization and (b) corresponding to a decrease in potential and increase in stored capacity. The colors in (a) correspond to each respective peak height for each diffraction peak, where blue and black are the lowest values, orange and red are the highest values, and green lies in between the two on a linear scale. (c) In region I, no lithiation occurs, while in region II, lithiation (blue) occurs at surface and grain boundaries. In region III, lithium reacts with silicon cores, with an almost complete reaction of silicon with lithium occurring in region IV. Reprinted with permission from Reference Reference Wang, Wu, Key, Yang, Grey, Zhu and Graetz19. © 2013 Wiley. Note: Q, magnitude of the scattering vector of scattered x-rays in reciprocal space, in Å^–1.

In operando characterization—Spatially resolved studies

Another key advantage of using HEXs to probe batteries in operando is the ability to penetrate through large amounts of material and enable the interior volume and bulk response of the battery to be examined. The distribution of chemical phases and electrochemical responses within a cell can be mapped with high spatial and temporal resolution. Techniques for this purpose include focused, monochromatic x-ray diffraction combined with phase-contrast imaging and energy-dispersive x-ray diffraction (ED-XRD).^{Reference Chen, Ren, Qin, Wu, Ma, Ren, He, Sun and Amine7}

Shui et al.^{Reference Shui, Okasinski, Chen, Almer and Liu20} were able to spatially resolve and characterize the distribution of LiO₂ within the cathode region of a lithium-oxygen battery system, confirm the reversible formation of lithium peroxide during electrochemical cycling, and quantify the amount and size of LiO₂ crystallites co-located with catalysts within the cell using a microfocused 20-µm HEX beam. Using the same setup but with a larger (unfocused) beam (Figure 5), phase-contrast tomography was employed to identify the existence and microstructure of an irreversible side reaction involving LiOH.^{Reference Shui, Okasinski, Kenesei, Dobbs, Zhao, Almer and Liu21} The porous structure of LiOH at the interface between the Li metal and the separator served as a pathway for Li-ions to travel across the layer and did not immediately shut down the operation of the battery. The compositional profiles and microstructures measured through the depth of each cell using the microfocused beam method allowed researchers to further understand and optimize catalyst composition and loading, as well as the pathways for both reversible and irreversible reactions in the Li-oxygen battery system.

Figure 5.

Schematic of multimodal experiments on a cylindrical battery cell. In full-field imaging mode, a large beam (W × H = 2.2 mm × 1.2 mm, where W is width and H is height) illuminates the (a) layered region of the battery cell, shown schematically. Phase-contrast tomographic measurements are made such that a 3D reconstruction with 1-µm resolution is achieved. (b) A representative sinogram of a LiOH layer on a metallic Li anode.^{Reference Shui, Okasinski, Kenesei, Dobbs, Zhao, Almer and Liu21} In a focused microbeam mode, the sample is translated across a small vertical beam (20 µm) along a relevant pathway, such as the dashed yellow line shown in (b), indicating the regions that are sampled by the focused microbeam. The scattering of the focused microbeam from the sample at each sampled point creates a scattering pattern that is collected (c) on a 2D area detector, where blue or black represents low intensities and red corresponds to high intensities on a linear scale.^{Reference Glazer, Cho, Almer, Okasinski, Braun and Dunand18} By translating the sample relative to the beam and quantifying the phase composition using the collected patterns, it is possible to create a (d) profile of the crystallographic phases that are present, where the distribution of lithium peroxide is shown. In (d), a low amount of Li₂O₂ is present in the gas diffusion layer (GDL), while a high amount is measured in the catalyst layer, with the highest amount near the catalyst–separator interface.^{Reference Shui, Okasinski, Chen, Almer and Liu20} Additionally, the slightly larger primary crystal size observed in the GDL suggests that the O₂-rich environment present in the GDL favors the growth of larger grains. Note: Q, magnitude of the scattering vector of scattered x-rays in reciprocal space, in Å^–1; Q _x and Q _y, in-plane components of the scattering vector measured on the 2D area detector. WAXS refers to wide-angle x-ray scattering; i.p.c., image proportional counts.

In addition to microfocused monochromatic beams, ED-XRD can also be used to resolve the spatial distribution of crystalline materials in operating batteries. Using ED-XRD, scattering of polychromatic x-rays (e.g., Bremsstrahlung radiation from a high atomic number target material laboratory source, or a polychromatic synchrotron source) is measured at a fixed diffraction angle using an energy-dispersive x-ray detector.^{Reference Liss, Bartels, Schreyer and Clemens1} Using this method, a finite volume within a complex environment can be probed, eliminating the need for customized, special-format batteries. The ED-XRD method has been successfully used to map spatial distributions in operando in various battery formats, including coin cell (lithium/silver vanadium diphosphate),^{Reference Kirshenbaum, Bock, Lee, Zhong, Takeuchi, Marschilok and Takeuchi22} prismatic (lithium iron phosphate),^{Reference Paxton, Zhong and Tsakalakos23} and grid scale (sodium halide).^{Reference Rijssenbeek, Gao, Zhong, Croft, Jisrawi, Ignatov and Tsakalakos24} For example, nine different points within a prismatic cell were monitored during electrochemical (dis)charge, with different degrees of chemical conversion from LiFePO₄ to FePO₄ observed within the cell at different locations.^{Reference Paxton, Zhong and Tsakalakos23} Such observations can aid in understanding how inhomogeneity within a battery cell affects the overall behavior of the cell.

Future directions

While this article focused on several scattering-based HEX techniques employed to study batteries, there are several other existing or developing imaging modalities that promise to benefit research in this area. In addition to conventional absorption-based imaging, these include (1) scattering tomography, which can be used to image crystalline and amorphous phases in materials;^{Reference Leemreize, Almer, Stock and Birkedal25} (2) three-dimensional (3D) x-ray diffraction, to image individual grains within polycrystalline aggregates;^{Reference Poulsen, Nielsen, Lauridsen, Schmidt, Suter, Lienert, Margulies, Lorentzen and Jensen26} and (3) coherent-diffractive imaging, which Ulvestad et al.^{Reference Ulvestad, Cho, Harder, Kim, Dietze, Fohtung, Meng and Shpyrko27} recently used to study single particles of battery materials.

It is anticipated that these methods will be increasingly used, either separately or in combination, to study 3D dynamic processes in battery materials, such as local distortions in the crystalline lattice during charge–discharge cycling. Given the inherently heterogeneous nature of battery systems, such imaging can be of great use to materials researchers, to reveal, for example, how (de)lithiation processes proceed across and within grains and how these processes affect aging.

Worldwide, several synchrotron facilities are planning to upgrade to “diffraction limited” storage rings, using multibend achromat technologies.^{Reference Borland28,Reference Reich29} These new machines will greatly increase the brilliance and coherence of the sources, particularly in the HEX range. This will significantly enhance the imaging techniques described, allowing sensitivity to the level of individual nanoscale particles within their bulk aggregates and minute concentrations or transient phases, such as those occurring at solid-electrolyte interfaces. While providing high scientific and technological impact on its own, HEX data are also anticipated to provide computational scientists with critical information needed to validate and extend multiscale models. HEX tools promise to help researchers design batteries with increased lifetimes and performance well into the future.

Summary

HEXs offer a powerful suite of capabilities for probing battery materials. The penetrating power of HEXs facilitates the use of a wide variety of sample environments for in situ and in operando observations, including high-temperature (>375°C) and high-pressure (>10 MPa) synthesis reaction chambers and commercial battery formats. Increasingly, techniques are being combined, such as SAXS, WAXS, and imaging, which allow a range of length scales to be probed. In the case of battery materials, these techniques have been used to reveal the nucleation, crystallization, and growth processes present during various synthesis routes, such as the formation processes of a nickel-rich full-concentration gradient oxide cathode.^{Reference Liss, Bartels, Schreyer and Clemens1} The stability and sensitivity of battery materials can also be probed inside thermal experimental cells, enabling the root causes responsible for thermal instability to be elucidated and ascertaining the effectiveness of mitigation strategies.

The high temporal resolution of HEXs, with pattern collection times lower than 150 ms possible on two-dimensional area detectors, allows for the identification of transient cycling mechanisms associated with rate-dependent performance and quantification of phase transitions during cycling. The identification of transient mechanisms and rapid phase transformations was previously possible only using a narrow scattering angle window over longer time durations when using a typical single-point detector and laboratory source, with full-pattern collection requiring at least three to five orders of magnitude greater time than that for the HEX pattern collection on area detectors. The high temporal resolution applies not only to experiments that examine a small sample volume, but also to bulk battery systems, where a large quantity of material can be probed with a high degree of spatial resolution. HEXs allow for compositional mapping in operando using SAXS and WAXS techniques and can further the understanding of the relationship between electrode morphology and inhomogeneity with overall performance.

Further improvements in brilliance and coherence at advanced HEX sources will enable even finer, more detailed information to be collected about material behavior at higher spatiotemporal resolutions, including individual electrode particles and minute transient phases, representing exciting prospects for strengthening the linkage between experimental and computational studies and for advancing battery science.