Probing and quantifying cathode charge heterogeneity in Li ion batteries†

Yuxin Zhang‡ ^a, Zhijie Yang‡ ^a and Chixia Tian *^b
aDepartment of Chemistry, Virginia Tech, Blacksburg, VA 24061, USA
^bAcademy of Integrated Science, Virginia Tech, Blacksburg, VA 24061, USA. E-mail: ctian@vt.edu

First published on 30th August 2019

---

---

Yuxin Zhang

Yuxin Zhang is a PhD student in the Department of Chemistry at Virginia Tech. He obtained his Bachelor's degree from Tianjin University in 2018. His current research focuses on understanding the charge transfer mechanism and interfacial reaction in aqueous batteries, as well as the investigation of promising cathode materials in non-aqueous lithium-ion batteries.

Zhijie Yang

Zhijie Yang is currently a PhD student in Chemistry at Virginia Tech. He obtained his Bachelor's degree from South China University of Technology (SCUT) in 2018. His research interest is in Co-free cathode materials and understanding the interfacial chemistry in Li-ion batteries.

Chixia Tian

Dr Chixia Tian is currently a Collegiate Assistant Professor at the Academy of Integrated Science at Virginia Tech. She received her Bachelor's degree in Polymer Materials Science and Engineering from Beijing University of Chemical Technology (2009) and PhD degree in Chemistry from Colorado School of Mines (2015). She was a post-doctoral researcher (2015–2017) at Lawrence Berkeley National Laboratory before she joined Virginia Tech. Her current research focuses on gaining a fundamental understanding of the degradation mechanism of cathode materials in lithium ion batteries and nanomedicine.

---

1. Introduction

The large consumption of fossil fuels has resulted in severe environmental issues and energy infrastructure challenges. There is an urgent need to make a paradigm shift in the energy landscape in the 21^st century.¹ Renewable energy resources are inevitably becoming an important alternative and a potential requirement to make such a shift. However, renewable energy resources, such as solar and wind, are usually intermittent and not always available, which necessitates strong energy storage systems that can make renewable energy available whenever and wherever it is needed.^2,3 On the other hand, the transportation sector, with regard to energy consumption, needs technological innovations that can either significantly increase the gasoline fuel efficiency (only as a temporary solution) or favorably and permanently enable a 100% electrical solution. Creating such a large paradigm shift in the energy landscape will require reliable, low-cost, high energy density, long life, and safe energy storage solutions. Battery technologies represent a family of such solution that has profoundly changed our lives.^4–7

After nearly 30 years of its commercialization and intense fundamental research, lithium ion batteries have become the technology of choice for electric vehicles, and their dominant role will continue for years to come.^8,9 The current goals in the field include but are not limited to increasing energy density, reducing cost, improving cycle and calendar life, and eliminating safety concerns. Achieving these goals involves efforts to improve the performance of battery electrodes and electrolytes.^10,11 Besides the new materials development, improving the existing materials is one of the practical pathways for achieving the near-term performance goals. Beyond the active materials, tremendous efforts have been made to optimize electrode formulation and fabrication,^12,13 improve cell integration and battery management,^14,15 and obtain a fundamental understanding of individual cell components and their interplay in the cell operating environment.¹¹ Collectively, these synergistic efforts have reduced the cost, improved the safety and energy density, and prolonged the cycle and calendar life of lithium ion batteries. As the lithium ion battery technology is on the verge of creating revolutionary changes in our transportation and energy landscape, it calls for more scientific and engineering innovations that can enable such a paradigm shift. Given the complex nature of a lithium ion battery, which was considered as a “black box”, elucidating how individual components perform their functionalities and how crosstalk takes place between the cathode and the anode in a battery is nontrivial.

Researchers were able to identify the most critical challenges facing lithium ion batteries.^8,16,17 These challenges include, but are not limited to, developing cathode materials with higher energy density to match the graphite anode,^18,19 increasing the electrolyte stability at high voltage,^10,11 reducing the cost not only on the materials side (reducing the cobalt amount)²⁰ but also on the manufacturing side (introducing new manufacturing processes),¹³ improving the safety by developing new electrolytes to replace the flammable liquid electrolytes that are currently widely adopted,^21,22 and recycling lithium ion batteries.^23,24 No matter which challenge is being tackled, it is inevitably associated with one of the three indispensable components of lithium ion batteries, namely the cathode, anode and electrolyte. Among the indispensable components, the cathode is the primary factor that determines the ultimate cell energy density.^18,19,25 The currently commercialized cathode materials are based on intercalation chemistry, where there are specific lattice sites for intercalation and de-intercalation of lithium ions. These cathode materials can be single crystal particles or polycrystalline particles that consist of many primary grains. In most cases, these primary grains are single crystals. Upon charging and discharging, there are redox reactions in specific ions in these particles. Depending on the type of material, phase transformations may take place concurrently with the redox reactions.^26–28 For other materials, the structural transformation is minor and thus the reaction follows a solid solution pathway.^29,30 No matter what reaction mechanism dominates in the particles, either phase transformation or solid solution, intriguing questions arise as to how a redox reaction initiates and propagates inside a particle, how redox reactions propagate inside an electrode, whether the ionic and electronic conducting pathways vary depending on the electrochemical protocol and/or electrode formulation, and how microscopic redox reactions determine the macroscale battery performance. Answering these questions would require a careful assessment of an important topic in the field: the state-of-charge heterogeneity (SOC) at multiple length scales in battery electrodes, from single particles, to multiple particles, and to battery electrodes.

The SOC is a crucial yet challenging to estimate macroscopic indicator of the state-of-health of a battery and is a key parameter used in the battery management system.³¹ Several methods have been proposed to estimate the SOC on the cell level,^32–35 average-electrode level,^36–38 and particle level^39–42 using experiments and simulations. The development of methods to accurately measure the SOC is critical, but the SOC heterogeneity has been a big obstacle for researchers to further improve cell performance and safety. The SOC heterogeneity can potentially impede the full activation of active materials and hence limit the energy density. Furthermore, the SOC heterogeneity could lead to local overcharge or overdischarge, which becomes a safety concern since oxygen has a tendency to be released from the material at high SOC, especially in the popular layered oxide cathodes.^43–46 Nonuniform SOC can also cause local stress, which facilitates crack formation and limits cell performance. Importantly, the crack formation increases the specific surface area of electrode particles and thus leads to more electrode–electrolyte side reactions.^47,48 With advanced characterization techniques that have mostly become available in the last 15 years, extensive SOC distribution studies at different length scales, from single particles to multiple particles, at the electrode level, and even at the cell level, have been reported.^{44–46,49–51} Many of these studies were enabled by synchrotron X-ray spectroscopic, imaging, and scattering techniques. These techniques can provide incisive diagnostics for battery chemistries, often times under operating and nondestructive conditions.

Improving charge homogeneity can increase the utilization of active materials, thus enhancing the energy density of batteries. There has been a large volume of studies in this area. However, to the best of our knowledge, the field currently lacks a comprehensive review to summarize the progress to further advance the field. Our review paper will focus on the charge distribution of lithium ion battery cathodes at multiple length scales ranging from single crystal particles to polycrystalline particles and composite electrodes (Fig. 1). Since different cathode materials follow distinct ion reaction mechanisms upon lithiation/delithiation, we will provide an in-depth discussion for each of them to generate a comprehensive scheme for the mechanism underlying the charge distribution heterogeneity. Our goal is also to provide some insights into establishing the relationship between the microscopic properties, macroscopic behaviors, and electrochemical performance.

2. Particle level heterogeneity

In the past decade, the rapid development of advanced synchrotron and electron techniques has enabled an improved understanding of charging behaviors and SOC distribution in battery particles, electrodes, and cells. These techniques make use of different interaction mechanisms between the X-ray/electron and battery particles, giving rise to a suite of advanced spectroscopic, imaging, and scattering techniques that are sensitive to phenomena at different length scales, from large-scale commercial batteries down to the atomic scale. Fig. 2a illustrates some of these techniques that are widely used in the literature.⁵²

As the charging and discharging take place in a cathode, lithium ions undergo deintercalation and intercalation, respectively. The corresponding changes are the redox potential and/or phases of the active cathode materials. When it comes to phase changes, X-ray powder diffraction (XRD) is a powerful tool that has been widely used, both ex situ and in situ, to identify different phases at different operation stages of a battery. XRD can be used to identify the lattice parameters of different phases and quantify the content of each phase.⁵³ With the advanced synchrotron based XRD and in situ technique, the temporal resolution has been significantly improved to allow researchers to capture the intermediate or metastable phases upon charging and discharging, which yields a more accurate conclusion about the actual phase transformation mechanism during electrochemical reactions rather than at the relaxed state in typical ex situ measurements. However, XRD is a bulk characterization technique that represents the ensemble-averaged information of many particles within the materials or composite electrodes.

The investigation of particle-level charge distribution on the nano- and micro-scale is essential to identify the underlying mechanisms that contribute to the heterogeneous SOC phenomena. The SOC heterogeneity within particles is directly related to the interior phase transformation. Different phases usually represent different oxidation states of transition metals and thus different lithium concentrations. Therefore, phase variation or phase separation is regarded as an indicator of charge distribution. In order to completely understand the complicated SOC heterogeneity on the particle level, techniques that are capable of capturing spatially resolved phase transformations and charge distribution are critical. Moreover, in situ and operando characterizations with superior temporal sensitivity are necessary to enable rapid acquisition of transient products under non-equilibrium conditions.

Transmission X-ray microscopy (TXM) and scanning transmission X-ray microscopy (STXM), in combination with X-ray absorption spectroscopy (XAS), are widely used to probe the oxidation states with tens of nanometers resolution and thus indicate the SOC heterogeneity at the single particle level.^42,46,54 Moreover, depending on the field of view and particle size, multiple particles can also be investigated simultaneously. TXM is a full-field technique, which means no movement of the sample during the imaging process. As shown in Fig. 2b and c, both nanoscale and microscale microscopy can be used for imaging. However, if the signal is collected pixel by pixel through sample rastering, it is called STXM, allowing for a flexible field of view. A more comprehensive comparison between the TXM and STXM techniques can be found in other review papers.^52,53 There are other powerful tools, such as Raman mapping and electron microscopy (e.g., scanning electron microscopy (SEM), TEM, and scanning transmission electron microscopy (STEM)). They can also be used to characterize charge distribution at the single particle or multiple particle level. We will introduce these techniques individually with specific examples throughout the review.

In the past few years, the SOC heterogeneity has been observed in various materials, including cathode and anode materials. In this article, we mainly focus on three specific types of cathode materials: olivine materials (e.g., LiFePO₄), layered materials (e.g., LiCoO₂, LiNi_1−x−yMn_xCo_yO₂, and LiNi_1−x−yCo_xAl_yO₂), and spinel materials (e.g., LiNi_0.5Mn_1.5O₄). These materials are either currently commercialized or have shown great potential for practical applications. Meanwhile, in order to gain a more fundamental understanding of the origin of the SOC heterogeneity at the particle level, the intrinsic properties of these materials, such as the phase transformation mechanism, lithium ion pathway, and particle size, are reviewed.

2.1. LiFePO₄ (LFP)

LFP attracts considerable attention due to its non-toxicity, low cost of raw materials, and high thermal stability.^55–57 Great progress has been made in understanding the phase transformation behavior and charge distribution of LFP upon charging and discharging. It was believed that LFP undergoes a first-order transformation without intermediate phases, where LiFePO₄ (LFP) converts to the FePO₄ (FP) phase when it is completely charged and reverts to the original phase (LFP) when it is completely discharged.^58–60 Recently, the development of more advanced characterization techniques has enabled the investigation of phase changes under in situ and/or operando conditions, actualizing the observation of intermediate products during electrochemical processes. Based on these studies, solid solution behavior is observed, which demonstrates that there are metastable phases upon cycling, and they ultimately disappear when the cell is rested. To date, there have been several review papers focusing on the LFP material and discussing the factors that impact the phase transformation behavior in this material.^26,28,61

Two different mechanisms, i.e., phase separation and solid solution, have been observed in LFP at the single particle level as well as at the electrode level. In a single particle, the phase separation mechanism indicates that a portion of the single particle is activated and undergoes a (de)lithiation process, and there is a phase boundary in the particle, separating the well-defined LFP and FP phases.⁵⁹ The solid solution mechanism proposes that the material experiences a transformation between the Li-rich (Li_xFePO₄) and Li-poor (Li_1−xFePO₄) phases without a clear phase boundary.^62,63 It is generally accepted that the solid solution exists in the non-equilibrium state while the phase separation behavior exists in the equilibrium state. This suggests that after relaxation, the metastable Li-rich and Li-poor phases relax to the stable LFP and FP phases. Correspondingly, one would expect that the SOC distribution also exhibits different patterns for the active and relaxed states, namely, a consecutive pattern and a distinct two-region pattern.

The phase separation mechanism in a single LFP particle was deduced by Goodenough and coworkers.⁵⁹ The authors proposed that the lithium intercalation was achieved by moving from the surface (LFP phase) to the bulk (FP phase), during which lithium ions would cross an interfacial Li_xFePO₄/Li_1−xFePO₄ layer.⁵⁹ In this “shrinking-core model” (Fig. 3a), the isotropic lithium ion insertion takes place along the radial pathway towards the center of the particle, indicating that the SOC in a single particle should be different at the core and the shell. Srinivasan et al. further elucidated the model and reported that from the beginning of discharge until the end, a LFP particle experienced the single phase–two phases–single phase process because the phase boundary gradually shrank as the intercalation proceeded (Fig. 3b).⁶⁴ However, further theoretical studies^67,68 showed that lithium ions in the LFP material would preferentially move along the b-axis rather than the a- or c-axis. Thus, the lithium diffusion in the LFP material is one-dimensional and there should be a preferable direction for the movement of lithium ions.⁶⁹ The “shrinking-core model” would create a large strain at the phase boundary, and there would be a thermodynamic and kinetic penalty.⁷⁰ Other models were proposed in the midst of resolving the phase separation mechanism.⁷¹ Allen et al. proposed the two-dimensional growth mechanism, which suggested that the lithium ion diffusion happened along the phase boundary (Fig. 3c).⁶⁵ This model was in excellent agreement with the microstructural evidence that was observed by Chen et al. with electron microscopy imaging of the FP/LFP phase boundary.⁷² Later, the “domino-cascade model” was proposed by Delmas et al.,⁶⁶ where the phase boundary movement along the direction that was perpendicular to the b-axis was believed to be faster compared to that in other directions (Fig. 3d), and the LFP–FP interface would immediately propagate to obtain a solely energy-favorable LFP or FP phase, robustly rationalizing some previous observations. Based on the “domino-cascade model”, the SOC heterogeneity should be more obvious in a certain direction compared to other directions.

Interestingly, a recent study combining X-ray absorption near edge structure (XANES) and TXM methods observed that lithium deintercalated along the preferable pathway only at the beginning of the charging process (Fig. 4). With further charging, the anisotropic pathway gradually transfers into the isotropic feature, which indicates that the phase boundary moves along many directions and shrinks uniformly. The reason could be that the extra driving force provided by the internal strain cannot be easily alleviated, and as the delithiation proceeds, the growing internal strain will force the phase propagation to move along multiple directions.⁷³ Thus, it is unrealistic to define an absolute isotropic or anisotropic phase transformation mechanism in LFP particles.

Nonetheless, well defined two-phase transformation models, either isotropic or anisotropic, cannot explain the sloping region appearing in the charge–discharge curves of the LFP material. A continuous non-equilibrium transformation, i.e., solid solution, would exist at the beginning and towards the end of cycling. In contrast to the two-phase transformation, there is no phase boundary in the solid solution behavior, which eliminates the significant energy barrier generated from the nucleation process. Compared to the SOC distribution of a phase separation system, the solid solution behavior theoretically exhibits a consecutive SOC variation in a single particle due to its continuous change of lithium content rather than distinct separation. In reality, however, many practical factors still influence the SOC distribution of a single particle with the solid solution behavior.⁷⁴ These factors include but are not limited to electrolyte infiltration, local electronic conducting pathways, and charging rates.^51,62,75

To fully understand the SOC distribution in the LFP material, the ion diffusion pathway in multiple particles should also be considered. By investigating the multiple particle system, a more comprehensive picture can be obtained to better understand how the SOC heterogeneity on the single particle level builds up and contributes to the SOC heterogeneity on the electrode level. Within a multiple particle system, two different lithium (de)intercalation pathways are proposed in the LFP material, namely, particle-by-particle (Fig. 5a and b) and concurrent intercalation (Fig. 5c and d).²⁶ Coexistence of fully charged and fully discharged particles and a small quantity of solid solution or two-phase separated particles can be detected if LFP undergoes particle-to-particle intercalation. On the other hand, each particle is activated and has similar lithium distributions if LFP undergoes concurrent intercalation. Chueh and coworkers revealed the particle-by-particle pathway in LFP particles by showing that most particles in electrodes were either fully charged or discharged, with only a few particles (2%) at the intermediate SOC, attributed to the fact that the time needed to finish charging a particle is much shorter than the time needed for nucleation.⁵⁴ Once the nucleation process is completed, the phase transformation will end immediately. Thus only the particles that undergo the nucleation process can be observed as active particles. This nucleation-limited kinetic model was supported by high resolution chemical analysis using energy-filtered TEM.⁷⁶ In the particle-by-particle intercalation pathway, even though the individual particles may have phase separation or solid solution reaction mechanisms, the bulk sensitive XRD cannot differentiate these mechanisms due to the limited number of active particles (exemplary XRD patterns in Fig. 5a and b). As a matter of fact, it is also challenging to use XRD to distinguish the differences between the particle-by-particle and concurrent pathways if they both undergo the phase separation mechanism. The concurrent solid solution reaction pathway, with the lowest SOC heterogeneity, has unique XRD diffraction patterns (Fig. 5d). Therefore, it is theoretically possible to identify this reaction pathway. However, it is challenging to make all the particles behave in the same fashion given the intricate environment in a composite electrode. The reaction pathway largely depends on the charging rate.²⁶ For example, a recent study discovered that a large current density favors the concurrent pathway.⁵⁰ Thus, the diffraction patterns and SOC distribution in multiple particles depend on both chemical and electrochemical factors; only by combining all the influencing elements in a specific case can we conclude which pathway is preferred and what the SOC distribution should be like.

2.2. Spinel materials

Spinel structured materials, such as LiMn₂O₄ and LiNi_0.5Mn_1.5O₄, have also been extensively studied as cathode materials in lithium-ion batteries because of their low cost, high rate capability and promising energy density.^25,77,78 In the spinel structure LiMn₂O₄, Li and Mn occupy the 8a tetrahedral and 16d octahedral sites of the cubic close-packed oxygen ions. There are four discrete phases in LiMn₂O₄ during cycling, namely, Li₂Mn₂O₄, LiMn₂O₄, Li_0.5Mn₂O₄ and Mn₂O₄. Fig. 6a–d clearly show that as the delithiation proceeds, Li₂Mn₂O₄ first exhibits an inhomogeneous charge distribution, which then changes to a homogeneous pattern when the voltage is close to 4 V, and finally it regains the inhomogeneous feature, which suggests a phase separation–solid solution–phase separation pathway occurring in the LiMn₂O₄ material.⁷⁹ In addition, the inhomogeneity is correlated not only with the intrinsic nature, but also with the crack formation upon charging. The crack can generate a mismatch between the local delithiation rate and overall charging rate, accelerating the SOC heterogeneity.

The Li₂Mn₂O₄ material faces severe capacity fading due to detrimental effects such as Jahn–Teller distortion, Mn²⁺ dissolution and development of micro-strains.^77,80–83 Therefore, substitution of Mn with other transition metals has been utilized to improve the structural stability and electrochemical performance. Among all the dopants, Ni substituted lithium manganese spinel, namely LiNi_0.5Mn_1.5O₄, has emerged as the most promising candidate and has been widely studied.^77,84

Similarly, three phases dominate the phase transformation process in the LiNi_0.5Mn_1.5O₄ material, namely, LiNi_0.5Mn_1.5O₄, Li_0.5Ni_0.5Mn_1.5O₄, and Ni_0.5Mn_1.5O₄. Through XRD measurement, three peaks, representing three different phases, can be detected during the charging and discharging process (Fig. 7a and b). LiNi_0.5Mn_1.5O₄ and Li_0.5Ni_0.5Mn_1.5O₄ phases coexist at lower potentials while Li_0.5Ni_0.5Mn_1.5O₄ and Ni_0.5Mn_1.5O₄ phases coexist at higher potentials.⁸⁴ A closer look into the redox chemistry in LiNi_0.5Mn_1.5O₄ reveals that not only are three different crystal structures involved, which can be detected by XRD, but the oxidation states of Ni are also different, being 2+ in LiNi_0.5Mn_1.5O₄, 3+ in Li_0.5Ni_0.5Mn_1.5O₄, and 4+ in Ni_0.5Mn_1.5O₄.^39,85,86 This redox chemistry difference offers researchers the opportunity to clearly visualize the oxidation state distribution, which is equivalent to the SOC distribution, in LiNi_0.5Mn_1.5O₄ materials with the help of the FF-TXM-XANES technique, similar to the LFP characterization we discussed earlier. As shown in Fig. 7c and d, both 2D and 3D Ni oxidation state maps illustrate the SOC heterogeneity in LiNi_0.5Mn_1.5O₄ during the cycling, and the 2D mapping image validates a three-phase concurrent phase transformation mechanism, which is different from the previous mechanism proposed by Hajime et al. which indicates that there are only two phases coexisting at various potentials. This discrepancy might be due to the difference in current rate and particle size.³⁹

2.3. Layered materials

Layered transition metal oxides (LiMO₂), which have high theoretical capacity (∼275 mA h g⁻¹) and a relatively high operation voltage (∼3.6 V versus Li⁺/Li), are widely used in commercial lithium-ion batteries.^29,77,87 The lithium intercalation pathway and the corresponding SOC heterogeneity in layered materials are not investigated as extensively as LFP since LiMO₂ materials mostly obey the solid solution behavior and possess better lithium diffusivity.⁴⁶ In such case, a relatively more homogeneous lithium distribution is expected. However, several studies, including ours, reported that the heterogeneous SOC distribution also exists in single-crystalline and polycrystalline layered materials under both equilibrium and non-equilibrium conditions.^{42,46,88–90}

LiCoO₂, with a theoretical capacity of 274 mA h g⁻¹, is one of the most successful cathode materials used in commercial batteries.^25,30,91,92 However, it can only provide half of its theoretical capacity (∼140 mA h g⁻¹) due to the structural instability at high states of charge when more than 0.5 Li is extracted from LiCoO₂.^25,93,94 A higher cutoff voltage (>4.2 V) is required to obtain a larger capacity, beyond 0.5 Li extraction. However, the overcharged particle (>0.5 Li extraction), induced by the high cutoff voltage, exhibits SOC heterogeneity and produces an uneven chemical distribution according to a recent study using STEM-electron energy-loss spectroscopy (EELS). Fig. 8a–f show the annular dark-field (ADF)-STEM images and the quantitative maps of the Li/Co ratio at the 40%, 60% and 100% charged states. Upon increasing the charging state, Li-poor areas are extended along the particle edges, leaving an inhomogeneous distribution of Li/Co ratios towards the end of charge.⁹⁵ Simultaneous lithium and oxygen extraction at high potentials is the main reason for this phenomenon. One may argue that the charge distribution inhomogeneity reported here is only at the several tens of nanometers scale. This is mainly because STEM-EELS provides high spatial resolution with the trade-off of limited experimental geometry. In this regard, TXM would be an ideal complementary tool to provide comprehensive information on the larger scale. Indeed, the SOC heterogeneity within individual LiCoO₂ particles was visualized in situ using TXM. Xu et al. discovered an inhomogeneous SOC distribution in a LiCoO₂ single particle after cycling at different current rates, as shown in Fig. 8g. A higher current rate causes a more severe SOC heterogeneity. The authors attributed this to the defects in LiCoO₂ that cause the rate-dependent nucleation process. It is worth pointing out that at each charged and discharged state, the SOC heterogeneity is always observed (green and red colored pixels in each single particle) regardless of the current rates. However, the extent (%) of SOC heterogeneity is largely dependent on the current densities.⁴¹

The high cost and environmentally unfriendly properties of Co and the aforementioned instability at high charge states have driven the efforts to identify alternative chemical compositions. The derivatives of LiCoO₂ attract much attention, in which Co is partially or completely substituted by earth-abundant transition metals, such as Ni and Mn. Among all the candidates, LiNi_1−x−yMn_xCo_yO₂ (NMC) has been at the spotlight of research.^29,87,96 By tuning the content of different transition metals, NMC materials offer different advantages to match the specific needs. For example, replacing Co with Ni improves the capacity by accessing Ni redox activity at lower voltages, while Mn offers higher thermal and structural stability.^96–98 Regardless of the NMC composition, based on the XRD studies, a clear solid solution behavior is determined upon cycling.^99,100 Gent et al. reported that a strongly nonuniform SOC distribution persisted within Li_1−xNi_1/3Co_1/3Mn_1/3O₂ (NMC333) secondary particles even after extensive relaxation (Fig. 9a).⁴⁶ This significant SOC heterogeneity remains even when different amounts of lithium are extracted (x_electrode = 0.30 and x_electrode = 0.60 in Fig. 9a). This was attributed to the anisotropic volume change in secondary particles, which induces internal stress and affects the (de)lithiation pathway. In another study, Tian et al. observed that an inhomogeneous charge distribution of NMC materials, in this case LiNi_0.6Co_0.2Mn_0.2O₂ (NMC622), exists both in chemically and electrochemically delithiated samples at the single particle level and among multiple particles (Fig. 9b–d).⁴² The charge heterogeneity in the NMC secondary particles can be attributed to multiple origins. In the electrochemical cycling, the electrical wiring, electrode porosity, and electrolyte penetration all contributed to the charge heterogeneity. In contrast, these origins did not exist in the chemical delithiation as it was carried out in a homogeneous suspension. In addition, the SOC heterogeneity is not transient in the layered materials such as the NMC materials shown in Fig. 9a. The heterogeneity could be observed even after relaxing the partially charged particles for 170 h.⁴⁶ This study was performed using a conventional electrode. The concern was that the local conducting pathways might also add an external factor that will influence the SOC distribution. Chemical delithiation is a widely used method to create a large quantity of materials for characterization. More importantly, the delithiation is done in solution typically with magnetic stirring. Therefore, it is fair to assume that different regions on the particle should not have limited access to the reaction medium. However, Tian et al. still observed the SOC heterogeneity using the chemical delithiation method, which suggested that the SOC heterogeneity is intrinsic to the polycrystalline NMC materials and was in agreement with the study done by Chueh and coworkers.⁴⁶

Such SOC heterogeneity phenomenon was also reported for the LiNi_0.8Co_0.15Al_0.05O₂ (NCA) material. Theoretically, NCA has a similar solid solution behavior to NMC materials. Studies also directly observed the SOC heterogeneity through the combination of FF-TXM and XANES. The electrochemical cycling profile and each step (steps I–VIII) stopped for Ni oxidation state mapping are shown in Fig. 10a. At each step, the same field of view is used and the particles that are under monitoring are shown in Fig. 10b. The Ni oxidation states are color coded for better visualization in Fig. 10c–k.⁸⁹ The red color indicates the lithiated state and the blue color indicates the delithiated state. The mixture of different colors at each step (each view) clearly demonstrates that SOC heterogeneity exists throughout the charging and discharging processes. When NCA was charged from 3.7 V to 4.2 V, the SOC varied from particle to particle. And as shown in Fig. 10k, when a large particle is closely investigated by mapping the Ni⁴⁺ (delithiated NCA), it clearly shows a ring-like delithiation profile, which is different from the random SOC distribution we observed in NMC materials. This was probably due to the slow lithium ion diffusion of the NCA material. This study was able to visualize and quantify the lithiation kinetics in NCA cathode materials with sub-particle resolution.

Even though TXM is a powerful technique to probe the SOC distribution of various materials, it is worth pointing out some other useful techniques that have also been used in constructing SOC maps on the particle level. Nanda et al. carried out micrometer-resolution Raman spectroscopic analysis to study the local SOC variation of NCA over a length scale of tens of micrometers within secondary particles at electrode surfaces and within the bulk of electrodes ex situ. The Raman scattering signal originates from the Raman active (A_1g + E_g) modes corresponding to oxygen vibrations in directions parallel (A_1g) or perpendicular (E_g) to the c-axis. For NCA materials, the E_g mode exhibits a band around 475 cm⁻¹, while the A_1g mode shows a band around 550 cm⁻¹. The values obtained from (A₄₇₅/A₅₅₀) × (I₄₇₅/I₅₅₀), where A and I are the peak area and intensity of the indicated bands, are used to semi-quantify the SOC state. Fig. 10l shows the heterogeneous SOC distribution of NCA across the electrode surface and through the edge. In terms of the overall electrode level, a higher oxidation state in the periphery of the particles was observed based on the quantitative analysis of Raman spectra (Fig. 10m).⁴⁵

Layered oxide chemistry offers a rich domain to manipulate the TM ratio in NMC. Different chemical compositions give rise to unique electrochemical and structural properties. Typically, a higher nickel content can deliver a higher capacity in a certain voltage window, which is due to the fact that nickel is the predominant redox center in NMC materials. However, nickel-rich compositions have their own limitations. Their surface structure is generally less stable than that of their nickel-poor counterparts. Furthermore, the phase transformation at high voltages becomes much more obvious for nickel-rich compositions. An extreme scenario is the LiNiO₂ material, which exhibits several plateau regions in the voltage profile. In general, each plateau represents a phase transformation event. During phase transformation, there will be a buildup of internal stress that can lead to more drastic accumulation of charge heterogeneity. To date, a complete comparison of charge heterogeneity between NMC materials with different chemical compositions has not been accomplished. A comparison study of this dimension is nontrivial, because one would need to consider the morphology and crystallographic orientation of grains in the polycrystalline NMC materials, which are also dependent on the chemical composition. The introduction of extra lithium ions can increase the practical capacity of layered materials (>250 mA h g⁻¹).^49,101,102 As a result, a series of Li-rich NMC materials were prepared with excess lithium ions located in the transition metal layer which could be extracted upon charging. Lithium–manganese rich NMCs (LMR-NMCs) are of particular interest and it is still debatable whether the pristine material is a mixture of nanodomains with two phases (i.e., LiMO₂ and Li₂MnO₃) or a solid solution.¹⁰¹ The goal of including this material here is to point out that the SOC heterogeneity also exists in this material.^49,103

3. Electrode level heterogeneity

In the previous section, we discussed the SOC heterogeneity on the single particle level and the multiple-particle system of different cathode materials. The discussion was based on the characterization methods with the SOC sensitivity at the atomic and nanoscale. However, the single particle level SOC heterogeneity cannot be isolated from the surrounding environments in composite electrodes.^45,46,89 For a few studies discussed in the previous single particle level section, the lithium was extracted using the chemical delithiation method.^39,42 The electrochemical processes are more practically relevant when the SOC heterogeneities are correlated with the battery performance. Therefore, we will focus on the SOC heterogeneity at the macroscopic length scale, that is the electrode level, in the following discussion. We will not categorize the electrode level SOC heterogeneity based on the type of cathode materials; instead we will frame our discussion based on the dominant factors that cause the macroscopic SOC heterogeneity.

During the charging or discharging process, an inhomogeneous distribution of the reaction and SOC at the electrode level has been observed in both the in-plane direction and in-depth direction with different types of cathode materials. These types of inhomogeneities can lead to the degradation of electrochemical performance with respect to capacity, energy density, and cycle life. In the following sections, both in-plane and in-depth direction inhomogeneities, as well as the factors resulting in the inhomogeneities will be discussed.

3.1. In-depth heterogeneity

In the lithium ion battery electrode fabrication process, one of the most common methods is to apply a slurry composed of the active material, conductive additive and binder on the current collector. The electrode thickness could vary from dozens of micrometers to hundreds of micrometers. To meet the high energy density demand, ideally, thicker electrodes need to be fabricated. However, it is important to understand whether all of the active material along the in-depth direction is fully activated as the electrodes become thicker. In an effort to investigate the SOC distribution along the in-depth direction, several groups have done research in this regard using different characterization tools.^{37,44,88,104,105}

In one of the examples, H. Murayama et al.⁸⁸ used in situ spectroscopic confocal XRD to monitor the lithium insertion process of NMC333 electrodes. Unlike the conventional XRD methods, a fixed detector with a confocal setup was used in this study to demonstrate the electrochemical reaction at different depths of the electrode (Fig. 11a). Continuous monochromatic X-rays were used as well to enhance the peak resolution. Since the interplanar distance d of the (113) planes changes during discharging and resting over time, the evolution of the intensity and position of the (113) peak with time can be utilized to estimate the SOC during these processes. As shown in the inset of Fig. 11b, the lithium content in the NMC333 material is roughly proportional to the d spacing within the experimental operating voltage window. To study the in-depth direction SOC (or lithium) distribution, in situ XRD was carried out at distances of 20 μm, 70 μm, and 120 μm from the surface of the electrode on the counter electrode side (Fig. 11c).

As shown in Fig. 11d–f, lithium insertion increases proportionally with time during discharging. The insertion rates can be estimated as 0.83, 0.71, and 0.33 h⁻¹ for the positions at 20 μm, 70 μm, and 120 μm depth, respectively. Compared to the overall discharging rate of 0.5C (=0.5 h⁻¹), the ion insertion rate at 20 μm is relatively higher while it is much lower at 120 μm. In other words, there is a lithium ion gradient forming from the counter electrode side (20 μm depth) to the current collector side (120 μm depth). However, after resting for tens of minutes, the lithium ion distribution at different depths reached homogeneity eventually. The authors proposed that the initial heterogeneous distribution is due to the larger lithium ion diffusion rate on the electrode in the vicinity of the electrolyte, while with sufficient lithium ion supply, the reaction inhomogeneity should be eliminated in this case by tuning porosity or other factors, which will be discussed in a later section.⁸⁸

Even though our review paper focuses on cathode materials and cathode electrodes, it is worth mentioning that the in-depth direction SOC inhomogeneity was also observed in graphite electrodes. S. J. Harris et al. created a model to demonstrate the in situ lithium spatial distribution over time, which helps to understand the charge and reaction distribution in the graphite electrode.¹⁰⁴ A half-cell composed of graphite and lithium foil has been made. As shown in Fig. 12a–d, different colors correspond to different states of lithiated graphite. The initial grey color in Fig. 12a represents the lowest Li content. In the following figures, blue corresponds to the state where graphite is lithiated at the LiC₁₈ concentration, called “dilute stage 2”. Further lithiation leads to the red color and it corresponds to the LiC₁₂ phase (stage 2). And finally, the gold color corresponds to the state where a fully lithiated state is reached with the LiC₆ phase (stage 1). Quantitatively, the boundary of red and gold regions represents roughly 90% lithiation (Li_0.9C₆). The optical micrographs show that Li diffuses gradually from the edge of the counter electrode side to the center and the lithiation process is nonuniform. To quantify the process, a log–log scale plot of the golden region width measured in pixels over time has been constructed (Fig. 12e). The experimental data points nearly fall on a straight line, which is described by the linear function in the figure. The authors assigned this process to the lithium diffusion based on the lithium insertion rate, although other processes might not be precluded. More discussion about the model can be found in the original reference. Note that the color change in this study only semi-quantitatively shows the concentration of inserted lithium; thus, more quantitative characterization methods are needed to further illustrate the SOC distribution in either graphite electrodes or other electrode materials.¹⁰⁴

3.2. In-plane heterogeneity

Besides the in-depth direction SOC distribution heterogeneity, such heterogeneity also exists along the in-plane direction. The non-uniform pressure applied to the electrode surface could be an origin of such in-plane heterogeneity. The in-plane variation of electrical resistance may also accelerate the SOC heterogeneity. The overcharged region in the electrode surface may become the electrode–electrolyte side reaction “hotspot”, resulting in local oxygen release from an oxide cathode.⁴³

The two-dimensional imaging mode of X-ray absorption fine structure (XAFS) is one of the best characterization methods to study the chemical valence states of transition metals in electrodes, where spatial resolution is not as critical as it is for single or multiple particle systems. In two consecutive studies reported by Katayama and co-workers, the SOC heterogeneity was observed for both LiMn₂O₄ and LFP cathodes. By mapping the chemical states of Mn (for LiMn₂O₄) and Fe (for LFP), they were able to color code a large area of the electrodes with the oxidation states of specific transition metals.^37,105Fig. 13a and b clearly show the heterogeneous mixing of different Mn oxidation states in a 50%-charged electrode, where blue and red pixels indicate the areas with the lower [Mn(III,IV)] and higher oxidation states [Mn(IV)], respectively. The yellow pixels represent regions that are intermediate between them.¹⁰⁵ This same characterization technique was applied for in situ experiments. Fig. 13c–h show the in situ XAFS chemical state mapping images of a LFP electrode during the first two successive cycles. The red pixels represent Fe(III) in the FP, in other words, the fully charged state. The blue pixels correspond to Fe(II) in LFP, indicating the fully discharged state. It is obvious that the distribution of the SOC and reaction is not uniform during charging and discharging processes although it becomes uniform in the fully charged/discharged state.³⁷ In both cases, it allows the mapping of chemical species over a wide area with a spatial resolution of 10 μm × 10 μm.

In addition to the phase transformation cathode materials (e.g., LiMn₂O₄ and LFP), LiCoO₂, with a solid solution behavior, also exhibits the in-plane non-uniform SOC distribution. T. Nishi et al. studied the SOC distribution inhomogeneity of LiCoO₂ along the in-plane direction during the charging and discharging process using in situ Raman imaging.¹⁰⁶ Raman spectroscopy is a powerful characterization method for studying the electrode surface properties because of its surface sensitivity and high signal resolution. The SOC mapping and the compositional distribution at the electrode surface were obtained from Raman spectroscopy. Fig. 14a and b show the in situ Raman mapping images of the tested LiCoO₂ region during discharging. There is 10% LiCoO₂ remaining charged after the electrode is fully discharged. After the second discharging process, the proportion of the charged region increased to 16%. The in situ Raman results clearly showed the inhomogeneous SOC distribution as charging or discharging progresses.¹⁰⁶

To further elucidate the origin of the heterogeneity, the correlation between the LCO A_1g peak position change and SOC distribution was determined (Fig. 14c and d) with the help of in situ Raman spectra. The A_1g mode in the LCO lattice corresponds to the Raman peak at 596 cm⁻¹. With increasing Li ion content in the LCO lattice, the frequency of the A_1g peak shifts to higher wavenumber, which makes it a good indicator of the SOC change. Comparing the 1^st cycle with the 2^nd cycle during both charging and discharging processes, the peak was broader in the 2^nd cycle, which indicates a larger range of the Li ion content in the lattice and a broader SOC distribution. This corresponds to high inhomogeneity of the local charge distribution in the electrode. The authors attributed the increasing SOC heterogeneity to the increasing electronic resistance between active material particles, which comes from SEI formation or mechanical stress.¹⁰⁶

A good example in Fig. 15a–c shows both the in-plane and in-depth SOC heterogeneities with the help of STXM.¹⁰⁷ The SOC distribution heterogeneity along the in-plane direction is random, and the in-depth distribution heterogeneity is reflected in the particle-by-particle level SOC heterogeneity, though no obvious change of the overall SOC of the selected regions has been observed. This could be attributed to the limited field of view with STXM as the scale bar shown in Fig. 15a–c is limited to 500 nm. In another study, the reaction (i.e., SOC distribution) of a thick electrode (∼148.6 mg cm⁻² loading) of LFP active materials was investigated with spatially resolved XRD computed tomography (XRD-CT) during charging and discharging (Fig. 15d). As shown in Fig. 15e, the reaction was accelerated at the electrode layers near the separator and the current collector during both charge and discharge, compared to layers at the center of the electrode. This clearly demonstrated that both ionic and electronic transport limit the reaction progress. Such in-depth SOC heterogeneity is quantified in Fig. 15f using lithium composition as the indicator for color coding. A maximum standard deviation of 50% in SOC heterogeneity is observed at the edge, while 20–40% variation was determined at the inner part of the electrode (Fig. 15g). Using the same characterization and quantification methods, the in-plane heterogeneity was also highlighted in Fig. 15h, which indicates a ∼10–20% maximum heterogeneity for planes (layers) that are closer to the separator and more than 30% maximum heterogeneity for layers that are further away from the separator. The authors also pointed out that the edge of the electrode (ring patterns in Fig. 15e) is another type of in-plane heterogeneity. This was most likely due to the fact that the edges are hindered from efficient electronic conductivity due to reduced compression of the cylindrical cell geometry.⁴⁴

In summary, the SOC distribution inhomogeneity along in-depth direction is usually attributed to the mismatch between the ion diffusion rate, electronic conductivity, and charging/discharging rate.^36,108,109 The side facing the electrolyte tends to have higher lithium ion concentration than the side facing the current collector during the lithiation process. But with sufficient lithium supply and enough diffusion time, the lithium ion distribution could reach homogeneity in the long range in the discussed cases. However, most batteries are operated under conditions far from equilibrium. Thus the SOC heterogeneity is most likely persistent throughout the life of a battery electrode. On the other hand, the in-plane SOC distribution during charging and discharging is influenced by a two-dimensional compositional distribution (carbon, voids, and active particles) and the uniformity of the applied pressure. A direct outcome of the SOC heterogeneity is to induce local overcharge or overdischarge, leading to a non-uniform utilization of active particles. It is an on-going effort to minimize the SOC heterogeneity through controlling materials properties, electrode formulation, electrochemical protocols, and thermal conditions, as discussed next.

4. Influencing factors

The lithium ion battery is a complex electrochemical system as there are redox reactions, lithiation/delithiation, and phase transformation in some cases happening in the system concurrently. Many factors have to be considered to understand the local chemical environment change. On the single particle level, we have detailed the intrinsic phase change behavior of the active materials. In addition, the particle size, shape and crystal orientation also contribute to the SOC heterogeneity and will be discussed in this section.

4.1. Particle-level factors

4.2. Electrode-level factors

We have thus far discussed the different key factors that determine the particle level SOC heterogeneity, where we categorize them into particle size, morphology, and surface facet termination. With all these factors fixed, the SOC distribution can still be altered because of the electrode fabrication process. Composite electrodes, consisting of active materials, electron conductive materials (e.g., carbon black), and binders, are inherently heterogeneous and complex. Moreover, in contact with electrolytic solution, the electrolyte distribution and diffusion add another factor into the system. Since the electrochemical reaction occurs preferentially in areas with lower resistance, SOC distribution heterogeneity can be expected within composite electrodes.¹⁰⁹

In a full cell, it has been reported that the positive electrode primarily contributes to the overall cell resistance increase.¹²⁵ In this study, a “reconstruction method” was applied to determine where the resistance increases most significantly. NCA and graphite were used as the cathode and anode materials, respectively. The cells were cycled both at 20 °C and 60 °C for hundreds of cycles. Since the cell that was cycled at 60 °C did not show a recognizable change in either capacity or resistance, 60 °C was chosen to generate a substantial resistance buildup for further investigation of the mechanism behind the increasing cell resistance. The authors disassembled the cell after cycling and re-constructed it with new electrode material that had not undergone any cycling. The resistance of the reconstructed cell with a cycled positive electrode and a pristine negative electrode increased drastically, while the one with a cycled negative electrode and a pristine positive electrode showed some increase but it was marginal (Fig. 19a). The same conclusion was reached when the authors used a “three electrode method” to test the resistance change. And further studies indicate that changes in the cathode particle morphology and the local structural and electronic states are the main contributors. Therefore, more attention will be given to positive electrodes in the following discussion.

On the electrode level, the inhomogeneous reaction and SOC heterogeneity discussed in Section 3 are mostly attributed to the spatial variation of electronic and ionic resistance. The electronic resistance could arise from insufficient contact between the active material particles, as well as between active materials and inactive materials (conductive additives and polymer binder). The ionic resistance exists in the bulk electrolyte and composite electrode. Furthermore, the inefficient Li ion diffusion and charge transfer between the electrode and electrolyte can also lead to the increase of ionic resistance. Fig. 19b illustrates the pathways of electrons and Li ions within a composite electrode and where they might encounter difficulties in providing sufficient conductivity.¹⁰⁹ A more detailed discussion about each factor that contributes to the inherent internal resistance was provided by Battaglia and coworkers.¹²⁷ They defined five different resistances: (1) electronic resistance of the electrode, R_e; (2) resistance to Li ion transport in the electrolyte to the surface of active material particles, R_s; (3) resistance to Li ion diffusion through the solid electrolyte interphase (SEI) film, R_SEI (in the case of the cathode, there is also resistance to a film or reconstruction layer called the cathode electrolyte interphase (CEI), R_CEI); (4) resistance to Li ion charge transfer at the electrode/electrolyte interface, R_ct; (5) resistance to diffusion of Li ions within the bulk electrode, R_diff. In the following section, the non-uniform distribution of electronic and ionic resistance and different factors that affect them are discussed in detail.

4.3. Post-treatment and electrochemical cycling factors

The fundamental understanding of both particle and electrode levels is reaching higher level precision because of the development of advanced characterization tools as well as the large amount of experimental data that can be adopted for theoretical simulations to build better models. All the products from these efforts can provide insights for researchers to further optimize the materials and tackle the engineering issues during electrode fabrication. However, there are other contributors that are just as important as the materials synthesis, and electrode preparation and optimization. They are electrochemical cycling conditions and other post-treatment methods (or external conditions such as temperature).

5. Conclusion and perspectives

Charge distribution heterogeneity prevails in most battery materials and electrodes. This review provides a comprehensive discussion of such heterogeneity at many length scales from the single particle to the composite electrode. In an ideal electrode, all battery particles would participate in delivering the capacity in the same fashion. However, due to the intrinsic properties of battery particles and the surrounding chemical environment, the charge heterogeneity occurs at the particle level. Often times, charge heterogeneity can be found within a small nanodomain in individual particles. For example, in primary particles of phase separation materials (e.g., LiFePO₄ and LiNi_0.5Mn_1.5O₄), regions of different charged states are separated by phase boundaries. In polycrystalline NMC particles, the charge heterogeneity may be more severe due to the presence of grain boundaries that direct the redox propagation during charging and discharging. The electrode-level charge heterogeneity is created due to the unbalanced ion and electron transport in the composite electrodes. The charge heterogeneity is clearly impacted by the materials chemistry, electrode formulation, and electrochemical protocols. The degradation of electrochemical performance is an important issue to be considered during cell aging, and the performance degradation results from the charge distribution heterogeneity across the electrode to a large extent, i.e., some regions of the electrode deteriorate faster than others due to the charge difference. Electronic and ionic conductivity are among the most important parameters that would determine the charge distribution and electrochemical performance. In order to optimize the electronic and ionic conductivity, the factors discussed above could be tuned during the battery fabrication process. Calendering is an effective method to balance the electronic and ionic conductivity. A larger volumetric energy density will be obtained due to denser electrodes without changing the mass loading. However, pressing the electrode will also decrease the porosity and increase tortuosity, thus leading to a lower specific active area and less efficient lithium ion migration. Appropriate calendering is necessary to decrease the contact resistance, while excessive calendering will lead to reduced porosity and greater tortuosity, which will increase the charge transfer resistance.

Fully understanding the charge heterogeneity of this dimension requires the use of many advanced diagnostic techniques. Table S1 in the ESI† summarizes the pros and cons of the techniques discussed in this manuscript. The great advancement of synchrotron X-ray spectroscopy, diffraction, and imaging techniques has offered unprecedented insights into the heterogeneity as the battery field invests in fast charging batteries. The operando probing of charge heterogeneity under fast charging conditions represents a large knowledge and technical gap. A clear disadvantage of the synchrotron techniques, especially the imaging-based ones, is the low data acquisition speed. Therefore, many studies discussed in this review rely on the ex situ measurements. On the other hand, imaging typically results in large datasets, offering an opportunity to apply state-of-the-art data analysis methods, such as machine learning based approaches. We expect that more advanced techniques will be available in the foreseeable future. Finally, there are challenges associated with establishing the relationship between charge distribution and electrochemical performance. Clearly, the challenges are multifaceted and future studies should focus on eliminating or mitigating the charge heterogeneity, which may offer a path towards increasing the battery energy density and cycle life.

The charge distribution may show high dependence on the battery operating conditions. To date, most Li ion batteries have been developed for room temperature operation, including those in electric vehicles and grid energy storage. However, future applications of batteries may move beyond this comfort zone towards extreme conditions, such as low and high operating temperatures. The ion reaction kinetics in batteries are highly dependent on temperature. Furthermore, the degradation behavior of battery materials is also influenced by the temperature profile. Therefore, we believe that more efforts need to be made to understand how the charge distribution is affected by the operating temperature. Fast charging is another important topic that has received widespread attention in recent years. Performing high current density electrochemistry on batteries can completely alter the reaction pathways in battery materials. A typical example is the switch from biphasic intercalation to solid solution reaction in some cathode materials, such as LFP. In layered oxides, fast charging may lead to lithium loss in the cathode, which leaves behind an incompletely discharged cathode after cycling. Furthermore, fast charging also induces more rapid lattice parameter changes and the battery particles are much further away from thermodynamic equilibrium conditions. On the other hand, the thermal outcome during fast charging is not very well understood. All these factors will influence the charging kinetics of battery electrodes and need to be considered. Finally, there have been an increasing number of new lithium ion battery cathode materials being developed in recent years. A typical example is the Li-rich disordered rocksalt oxides. The ion reactions in these cathode materials do not follow the conventional intercalation reactions. Instead, there are three-dimensional Li ion percolating pathways, and in most cases, such pathways are randomized. Probing the charge heterogeneity in these new materials would be an interesting research topic. There have not been many studies investigating the full cell performance of the disordered rocksalt oxides. Currently, the US Department of Energy sponsors research to evaluate the likelihood of improving these materials for practical batteries. We expect that when the practical application of disordered rocksalt oxides is on the verge, more fundamental investigation of charge distribution at multiple length scales will be needed. Clearly, homogenizing the charge distribution and making individual particles contribute synchronously to battery performance are daunting challenges. We expect that advanced manufacturing techniques such as 3D printing will allow for targeted engineering of electrode architectures. In particular, if the spatial resolution of 3D printing can match the length scale of individual battery particles and additives, the technique may bring revolutionary advancements to the battery field. Certainly, the cost and advantages of 3D printing will need to be appealing enough to make a shift away from conventional electrode manufacturing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Academy of Integrated Science Startup Funds at Virginia Tech.

References

1. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303 CrossRef CAS PubMed .
2. O. Ellabban, H. Abu-Rub and F. Blaabjerg, Renewable Sustainable Energy Rev., 2014, 39, 748–764 CrossRef .
3. M. Beaudin, H. Zareipour, A. Schellenberglabe and W. Rosehart, Energy Sustainable Dev., 2010, 14, 302–314 CrossRef .
4. B. Scrosati and J. Garche, J. Power Sources, 2010, 195, 2419–2430 CrossRef CAS .
5. M. Winter, B. Barnett and K. Xu, Chem. Rev., 2018, 118, 11433–11456 CrossRef CAS PubMed .
6. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657 CrossRef CAS PubMed .
7. B. Dunn, H. Kamath and J.-M. Tarascon, Science, 2011, 334, 928–935 CrossRef CAS PubMed .
8. H. Li, Joule, 2019, 3, 911–914 CrossRef CAS .
9. E. Karden, S. Ploumen, B. Fricke, T. Miller and K. Snyder, J. Power Sources, 2007, 168, 2–11 CrossRef CAS .
10. K. Xu, Chem. Rev., 2004, 104, 4303–4417 CrossRef CAS PubMed .
11. K. Xu, Chem. Rev., 2014, 114, 11503–11618 CrossRef CAS PubMed .
12. G. Liu, H. Zheng, S. Kim, Y. Deng, A. M. Minor, X. Song and V. S. Battaglia, J. Electrochem. Soc., 2008, 155, A887 CrossRef CAS .
13. B. Ludwig, Z. Zheng, W. Shou, Y. Wang and H. Pan, Sci. Rep., 2016, 6, 23150 CrossRef CAS PubMed .
14. L. Lu, X. Han, J. Li, J. Hua and M. Ouyang, J. Power Sources, 2013, 226, 272–288 CrossRef CAS .
15. S. S. Lee, T. H. Kim, S. J. Hu, W. W. Cai and J. A. Abell, in ASME 2010 International Manufacturing Science and Engineering Conference, Vol. 1, ASME, 2010, pp. 541–549 Search PubMed .
16. J.-M. Tarascon and M. Armand, Nature, 2001, 414, 359–367 CrossRef CAS PubMed .
17. J. B. Goodenough and K.-S. Park, J. Am. Chem. Soc., 2013, 135, 1167–1176 CrossRef CAS PubMed .
18. J. Xu, S. Dou, H. Liu and L. Dai, Nano Energy, 2013, 2, 439–442 CrossRef CAS .
19. J. W. Fergus, J. Power Sources, 2010, 195, 939–954 CrossRef CAS .
20. Y.-K. Sun, D.-J. Lee, Y. J. Lee, Z. Chen and S.-T. Myung, ACS Appl. Mater. Interfaces, 2013, 5, 11434–11440 CrossRef CAS PubMed .
21. A. Manthiram, X. Yu and S. Wang, Nat. Rev. Mater., 2017, 2, 16103 CrossRef CAS .
22. E. Quartarone and P. Mustarelli, Chem. Soc. Rev., 2011, 40, 2525 RSC .
23. J. Xu, H. R. Thomas, R. W. Francis, K. R. Lum, J. Wang and B. Liang, J. Power Sources, 2008, 177, 512–527 CrossRef CAS .
24. X. Zeng, J. Li and N. Singh, Crit. Rev. Environ. Sci. Technol., 2014, 44, 1129–1165 CrossRef CAS .
25. M. S. Whittingham, Chem. Rev., 2004, 104, 4271–4301 CrossRef CAS PubMed .
26. Y. Li, Solid State Ionics, 2018, 323, 142–150 CrossRef CAS .
27. H. Komatsu, H. Arai, Y. Koyama, K. Sato, T. Kato, R. Yoshida, H. Murayama, I. Takahashi, Y. Orikasa, K. Fukuda, T. Hirayama, Y. Ikuhara, Y. Ukyo, Y. Uchimoto and Z. Ogumi, Adv. Energy Mater., 2015, 5, 1500638 CrossRef .
28. R. Malik, A. Abdellahi and G. Ceder, J. Electrochem. Soc., 2013, 160, A3179–A3197 CrossRef CAS .
29. W. Liu, P. Oh, X. Liu, M. Lee, W. Cho, S. Chae, Y. Kim and J. Cho, Angew. Chem., Int. Ed., 2015, 54, 4440–4457 CrossRef CAS PubMed .
30. K. Ozawa, Solid State Ionics, 1994, 69, 212–221 CrossRef CAS .
31. K. W. E. Cheng, B. P. Divakar, H. Wu, K. Ding and H. F. Ho, IEEE Trans. Veh. Technol., 2011, 60, 76–88 Search PubMed .
32. L. Cai, K. An, Z. Feng, C. Liang and S. J. Harris, J. Power Sources, 2013, 236, 163–168 CrossRef CAS .
33. S. Piller, M. Perrin and A. Jossen, J. Power Sources, 2001, 96, 113–120 CrossRef CAS .
34. S. Santhanagopalan and R. E. White, J. Power Sources, 2006, 161, 1346–1355 CrossRef CAS .
35. S. Lee, J. Kim, J. Lee and B. H. Cho, J. Power Sources, 2008, 185, 1367–1373 CrossRef CAS .
36. J. Liu, M. Kunz, K. Chen, N. Tamura and T. J. Richardson, J. Phys. Chem. Lett., 2010, 1, 2120–2123 CrossRef CAS .
37. M. Katayama, K. Sumiwaka, R. Miyahara, H. Yamashige, H. Arai, Y. Uchimoto, T. Ohta, Y. Inada and Z. Ogumi, J. Power Sources, 2014, 269, 994–999 CrossRef CAS .
38. T. Danner, M. Singh, S. Hein, J. Kaiser, H. Hahn and A. Latz, J. Power Sources, 2016, 334, 191–201 CrossRef CAS .
39. S. Kuppan, Y. Xu, Y. Liu and G. Chen, Nat. Commun., 2017, 8, 14309 CrossRef CAS PubMed .
40. Y. Li, J. N. Weker, W. E. Gent, D. N. Mueller, J. Lim, D. A. Cogswell, T. Tyliszczak and W. C. Chueh, Adv. Funct. Mater., 2015, 25, 3677–3687 CrossRef CAS .
41. Y. Xu, E. Hu, K. Zhang, X. Wang, V. Borzenets, Z. Sun, P. Pianetta, X. Yu, Y. Liu and X.-Q. Yang, ACS Energy Lett., 2017, 2, 1240–1245 CrossRef .
42. C. Tian, Y. Xu, D. Nordlund, F. Lin, J. Liu, Z. Sun, Y. Liu and M. Doeff, Joule, 2018, 2, 464–477 CrossRef CAS .
43. F. Lin, I. M. Markus, D. Nordlund, T.-C. Weng, M. D. Asta, H. L. Xin and M. M. Doeff, Nat. Commun., 2014, 5, 3529 CrossRef PubMed .
44. H. Liu, S. Kazemiabnavi, A. Grenier, G. Vaughan, M. Di Michiel, B. J. Polzin, K. Thornton, K. W. Chapman and P. J. Chupas, ACS Appl. Mater. Interfaces, 2019, 11, 18386–18394 CrossRef CAS PubMed .
45. J. Nanda, J. Remillard, A. O'Neill, D. Bernardi, T. Ro, K. E. Nietering, J. Go and T. J. Miller, Adv. Funct. Mater., 2011, 21, 3282–3290 CrossRef CAS .
46. W. E. Gent, Y. Li, S. Ahn, J. Lim, Y. Liu, A. M. Wise, C. B. Gopal, D. N. Mueller, R. Davis, J. N. Weker, J.-H. Park, S.-K. Doo and W. C. Chueh, Adv. Mater., 2016, 28, 6631–6638 CrossRef CAS PubMed .
47. S. Xia, L. Mu, Z. Xu, J. Wang, C. Wei, L. Liu, P. Pianetta, K. Zhao, X. Yu and F. Lin, Nano energy, 2018, 53, 753–762 CrossRef CAS .
48. Z. Xu, M. M. Rahman, L. Mu, Y. Liu and F. Lin, J. Mater. Chem. A, 2018, 6, 21859–21884 RSC .
49. F. Yang, Y. Liu, S. K. Martha, Z. Wu, J. C. Andrews, G. E. Ice, P. Pianetta and J. Nanda, Nano Lett., 2014, 14, 4334–4341 CrossRef CAS PubMed .
50. Y. Li, F. El Gabaly, T. R. Ferguson, R. B. Smith, N. C. Bartelt, J. D. Sugar, K. R. Fenton, D. A. Cogswell, A. L. D. Kilcoyne, T. Tyliszczak, M. Z. Bazant and W. C. Chueh, Nat. Mater., 2014, 13, 1149 CrossRef CAS PubMed .
51. H. Liu, F. C. Strobridge, O. J. Borkiewicz, K. M. Wiaderek, K. W. Chapman, P. J. Chupas and C. P. Grey, Science, 2014, 344, 1252817 CrossRef PubMed .
52. J. Nelson Weker and M. F. Toney, Adv. Funct. Mater., 2015, 25, 1622–1637 CrossRef CAS .
53. F. Lin, Y. Liu, X. Yu, L. Cheng, A. Singer, O. G. Shpyrko, H. L. Xin, N. Tamura, C. Tian and T.-C. Weng, Chem. Rev., 2017, 117, 13123–13186 CrossRef CAS PubMed .
54. W. C. Chueh, F. El Gabaly, J. D. Sugar, N. C. Bartelt, A. H. McDaniel, K. R. Fenton, K. R. Zavadil, T. Tyliszczak, W. Lai and K. F. McCarty, Nano Lett., 2013, 13, 866–872 CrossRef CAS PubMed .
55. A. Yamada, S. C. Chung and K. Hinokuma, J. Electrochem. Soc., 2001, 148, A224 CrossRef CAS .
56. S.-Y. Chung, J. T. Bloking and Y.-M. Chiang, Nat. Mater., 2002, 1, 123–128 CrossRef CAS PubMed .
57. H. Huang, S.-C. Yin and L. F. Nazar, Electrochem. Solid-State Lett., 2001, 4, A170 CrossRef CAS .
58. L. Hong, L. Li, Y.-K. Chen-Wiegart, J. Wang, K. Xiang, L. Gan, W. Li, F. Meng, F. Wang, J. Wang, Y.-M. Chiang, S. Jin and M. Tang, Nat. Commun., 2017, 8, 1194 CrossRef PubMed .
59. A. K. Padhi, K. S. Nanjundaswamy and J. B. Goodenough, J. Electrochem. Soc., 1997, 144, 1188–1194 CrossRef CAS .
60. B. Kang and G. Ceder, Nature, 2009, 458, 190–193 CrossRef CAS PubMed .
61. W.-J. Zhang, J. Power Sources, 2011, 196, 2962–2970 CrossRef CAS .
62. J. Niu, A. Kushima, X. Qian, L. Qi, K. Xiang, Y.-M. Chiang and J. Li, Nano Lett., 2014, 14, 4005–4010 CrossRef CAS PubMed .
63. C. Delacourt, P. Poizot, J.-M. Tarascon and C. Masquelier, Nat. Mater., 2005, 4, 254–260 CrossRef CAS .
64. V. Srinivasan and J. Newman, J. Electrochem. Soc., 2004, 151, A1517–A1529 CrossRef CAS .
65. J. L. Allen, T. R. Jow and J. Wolfenstine, J. Solid State Electrochem., 2008, 12, 1031–1033 CrossRef CAS .
66. C. Delmas, M. Maccario, L. Croguennec, F. Le Cras and F. Weill, Nat. Mater., 2008, 7, 665–671 CrossRef CAS PubMed .
67. D. Morgan, A. Van der Ven and G. Ceder, Electrochem. Solid-State Lett., 2004, 7, A30 CrossRef CAS .
68. M. Saiful Islam, D. J. Driscoll, A. Craig, A. J. Fisher and P. R. Slater, Chem. Mater., 2005, 17, 5085–5092 CrossRef .
69. M. S. Whittingham, Y. Song, S. Lutta, P. Y. Zavalij and N. A. Chernova, J. Mater. Chem., 2005, 15, 3362–3379 RSC .
70. N. Meethong, H. Huang, S. A. Speakman, W. C. Carter and Y. Chiang, Adv. Funct. Mater., 2007, 17, 1115–1123 CrossRef CAS .
71. G. Brunetti, D. Robert, P. Bayle-Guillemaud, J. L. Rouvière, E. F. Rauch, J. F. Martin, J. F. Colin, F. Bertin and C. Cayron, Chem. Mater., 2011, 23, 4515–4524 CrossRef CAS .
72. G. Chen, X. Song and T. J. Richardson, Electrochem. Solid-State Lett., 2006, 9, A295–A298 CrossRef CAS .
73. J. Wang, Y. K. Chen-Wiegart, C. Eng, Q. Shen and J. Wang, Nat. Commun., 2016, 7, 12372 CrossRef CAS PubMed .
74. G. Chen, X. Song and T. J. Richardson, J. Electrochem. Soc., 2007, 154, A627–A632 CrossRef CAS .
75. Y. Orikasa, T. Maeda, Y. Koyama, H. Murayama, K. Fukuda, H. Tanida, H. Arai, E. Matsubara, Y. Uchimoto and Z. Ogumi, J. Am. Chem. Soc., 2013, 135, 5497–5500 CrossRef CAS PubMed .
76. J. D. Sugar, F. El Gabaly, W. C. Chueh, K. R. Fenton, T. Tyliszczak, P. G. Kotula and N. C. Bartelt, J. Power Sources, 2014, 246, 512–521 CrossRef CAS .
77. B. Xu, D. Qian, Z. Wang and Y. S. Meng, Mater. Sci. Eng., R, 2012, 73, 51–65 CrossRef CAS .
78. M. M. Thackeray, W. I. F. David, P. G. Bruce and J. B. Goodenough, Mater. Res. Bull., 1983, 18, 461–472 CrossRef CAS .
79. Y. Yu, C. Kim, Y. Liu, A. Van der Ven, Y. S. Meng, R. Kostecki and J. Cabana, Adv. Energy Mater., 2015, 5, 1402040 CrossRef .
80. H. Xia, Z. Luo and J. Xie, Prog. Nat. Sci.: Mater. Int., 2012, 22, 572–584 CrossRef .
81. M. M. Thackeray, Prog. Solid State Chem., 1997, 25, 1–71 CrossRef CAS .
82. M. Yonemura, A. Yamada, H. Kobayashi, M. Tabuchi, T. Kamiyama, Y. Kawamoto and R. Kanno, J. Mater. Chem., 2004, 14, 1948 RSC .
83. N. Leifer, F. Schipper, E. M. Erickson, C. Ghanty, M. Talianker, J. Grinblat, C. M. Julien, B. Markovsky and D. Aurbach, J. Phys. Chem. C, 2017, 121, 9120–9130 CrossRef CAS .
84. H. Arai, K. Sato, Y. Orikasa, H. Murayama, I. Takahashi, Y. Koyama, Y. Uchimoto and Z. Ogumi, J. Mater. Chem. A, 2013, 1, 10442–10449 RSC .
85. S. Mukerjee, X. Q. Yang, X. Sun, S. J. Lee, J. McBreen and Y. Ein-Eli, Electrochim. Acta, 2004, 49, 3373–3382 CrossRef CAS .
86. K. Ariyoshi, Y. Iwakoshi, N. Nakayama and T. Ohzuku, J. Electrochem. Soc., 2004, 151, A296–A303 CrossRef CAS .
87. J. Xu, F. Lin, M. M. Doeff and W. Tong, J. Mater. Chem. A, 2017, 5, 874–901 RSC .
88. H. Murayama, K. Kitada, K. Fukuda, A. Mitsui, K. Ohara, H. Arai, Y. Uchimoto, Z. Ogumi and E. Matsubara, J. Phys. Chem. C, 2014, 118, 20750–20755 CrossRef CAS .
89. L. Nowack, D. Grolimund, V. Samson, F. Marone and V. Wood, Sci. Rep., 2016, 6, 21479 CrossRef CAS PubMed .
90. Y. Zhou, J. Yue, E. Hu, H. Li, L. Gu, K. Nam, S. Bak, X. Yu, J. Liu and J. Bai, Adv. Energy Mater., 2016, 6, 1600597 CrossRef .
91. E.-D. Jeong, M.-S. Won and Y.-B. Shim, J. Power Sources, 1998, 70, 70–77 CrossRef CAS .
92. K. Mizushima, P. C. Jones, P. J. Wiseman and J. B. Goodenough, Solid State Ionics, 1981, 3–4, 171–174 CrossRef CAS .
93. D. Deng, Energy Sci. Eng., 2015, 3, 385–418 CrossRef .
94. V. Etacheri, R. Marom, R. Elazari, G. Salitra and D. Aurbach, Energy Environ. Sci., 2011, 4, 3243 RSC .
95. J. Kikkawa, S. Terada, A. Gunji, T. Nagai, K. Kurashima and K. Kimoto, J. Phys. Chem. C, 2015, 119, 15823–15830 CrossRef CAS .
96. A. Manthiram, B. Song and W. Li, Energy Storage Mater., 2017, 6, 125–139 CrossRef .
97. Y.-K. Sun, Z. Chen, H.-J. Noh, D.-J. Lee, H.-G. Jung, Y. Ren, S. Wang, C. S. Yoon, S.-T. Myung and K. Amine, Nat. Mater., 2012, 11, 942–947 CrossRef CAS PubMed .
98. H.-J. Noh, S. Youn, C. S. Yoon and Y.-K. Sun, J. Power Sources, 2013, 233, 121–130 CrossRef CAS .
99. M. Ma, N. A. Chernova, B. H. Toby, P. Y. Zavalij and M. S. Whittingham, J. Power Sources, 2007, 165, 517–534 CrossRef CAS .
100. K. Märker, P. J. Reeves, C. Xu, K. J. Griffith and C. P. Grey, Chem. Mater., 2019, 31, 2545–2554 CrossRef .
101. A. K. Shukla, Q. M. Ramasse, C. Ophus, H. Duncan, F. Hage and G. Chen, Nat. Commun., 2015, 6, 8711 CrossRef CAS PubMed .
102. A. K. Shukla, Q. M. Ramasse, C. Ophus, D. M. Kepaptsoglou, F. S. Hage, C. Gammer, C. Bowling, P. A. H. Gallegos and S. Venkatachalam, Energy Environ. Sci., 2018, 11, 830–840 RSC .
103. B. Xu, C. R. Fell, M. Chi and Y. S. Meng, Energy Environ. Sci., 2011, 4, 2223–2233 RSC .
104. S. J. Harris, A. Timmons, D. R. Baker and C. Monroe, Chem. Phys. Lett., 2010, 485, 265–274 CrossRef CAS .
105. M. Katayama, K. Sumiwaka, K. Hayashi, K. Ozutsumi, T. Ohta, Y. Inada and IUCr, J. Synchrotron Radiat., 2012, 19, 717–721 CrossRef CAS PubMed .
106. T. Nishi, H. Nakai and A. Kita, J. Electrochem. Soc., 2013, 160, A1785–A1788 CrossRef CAS .
107. W. C. Chueh, F. El Gabaly, J. D. Sugar, N. C. Bartelt, A. H. McDaniel, K. R. Fenton, K. R. Zavadil, T. Tyliszczak, W. Lai and K. F. McCarty, Nano Lett., 2013, 13, 866–872 CrossRef CAS PubMed .
108. T. Sasaki, C. Villevieille, Y. Takeuchi and P. Novák, Adv. Sci., 2015, 2, 1500083 CrossRef PubMed .
109. Y. Orikasa, Y. Gogyo, H. Yamashige, M. Katayama, K. Chen, T. Mori, K. Yamamoto, T. Masese, Y. Inada, T. Ohta, Z. Siroma, S. Kato, H. Kinoshita, H. Arai, Z. Ogumi and Y. Uchimoto, Sci. Rep., 2016, 6, 26382 CrossRef CAS PubMed .
110. Y. Hu, M. M. Doeff, R. Kostecki and R. Fiñones, J. Electrochem. Soc., 2004, 151, A1279 CrossRef CAS .
111. T. Drezen, N.-H. Kwon, P. Bowen, I. Teerlinck, M. Isono and I. Exnar, J. Power Sources, 2007, 174, 949–953 CrossRef CAS .
112. Y. Wang and G. Cao, Adv. Mater., 2008, 20, 2251–2269 CrossRef CAS .
113. C. Jiang, E. Hosono and H. Zhou, Nano Today, 2006, 1, 28–33 CrossRef .
114. J. Li, R. Shunmugasundaram, R. Doig and J. R. Dahn, Chem. Mater., 2015, 28, 162–171 CrossRef .
115. R. Dominko, M. Gaberš, J. Drofenik, M. Bele and S. Pejovnik, Electrochem. Solid-State Lett., 2001, 4, A187 CrossRef CAS .
116. B. Lung-Hao Hu, F.-Y. Wu, C.-T. Lin, A. N. Khlobystov and L.-J. Li, Nat. Commun., 2013, 4, 1687 CrossRef PubMed .
117. N. Meethong, H.-Y. S. Huang, W. C. Carter and Y.-M. Chiang, Electrochem. Solid-State Lett., 2007, 10, A134–A138 CrossRef CAS .
118. M. Gaberscek, R. Dominko and J. Jamnik, Electrochem. Commun., 2007, 9, 2778–2783 CrossRef CAS .
119. M. Wagemaker, F. M. Mulder and A. Van der Ven, Adv. Mater., 2009, 21, 2703–2709 CrossRef CAS .
120. A. Yamada, H. Koizumi, S. Nishimura, N. Sonoyama, R. Kanno, M. Yonemura, T. Nakamura and Y. Kobayashi, Nat. Mater., 2006, 5, 357 CrossRef CAS PubMed .
121. P. Gibot, M. Casas-Cabanas, L. Laffont, S. Levasseur, P. Carlach, S. Hamelet, J.-M. Tarascon and C. Masquelier, Nat. Mater., 2008, 7, 741 CrossRef CAS PubMed .
122. D. A. Cogswell and M. Z. Bazant, ACS Nano, 2012, 6, 2215–2225 CrossRef CAS PubMed .
123. K. Saravanan, P. Balaya, M. V Reddy, B. V. R. Chowdari and J. J. Vittal, Energy Environ. Sci., 2010, 3, 457–463 RSC .
124. K. R. Chemelewski, D. W. Shin, W. Li and A. Manthiram, J. Mater. Chem. A, 2013, 1, 3347–3354 RSC .
125. Y. Itou and Y. Ukyo, J. Power Sources, 2005, 146, 39–44 CrossRef CAS .
126. R. Dominko, M. Gaberscek, J. Drofenik, M. Bele, S. Pejovnik and J. Jamnik, J. Power Sources, 2003, 119–121, 770–773 CrossRef CAS .
127. H. Zheng, J. Li, X. Song, G. Liu and V. S. Battaglia, Electrochim. Acta, 2012, 71, 258–265 CrossRef CAS .
128. C. Heubner, A. Nickol, J. Seeba, S. Reuber, N. Junker, M. Wolter, M. Schneider and A. Michaelis, J. Power Sources, 2019, 419, 119–126 CrossRef CAS .
129. J. Jeong, H. Lee, J. Choi, M.-H. Ryou and Y. M. Lee, Electrochim. Acta, 2015, 154, 149–156 CrossRef CAS .
130. M. Doyle, J. Newman, A. S. Gozdz, C. N. Schmutz and J. Tarascon, J. Electrochem. Soc., 1996, 143, 1890 CrossRef .
131. D. Dees, E. Gunen, D. Abraham, A. Jansen and J. Prakash, J. Electrochem. Soc., 2008, 155, 603–613 CrossRef .
132. Z. Siroma, J. Hagiwara, K. Yasuda, M. Inaba and A. Tasaka, J. Electroanal. Chem., 2010, 648, 92–97 CrossRef CAS .
133. C. W. Wang, A. M. Sastry, K. A. Striebel and K. Zaghib, J. Electrochem. Soc., 2005, 152, A1001–A1010 CrossRef CAS .
134. C. Heubner, A. Nickol, J. Seeba, S. Reuber, N. Junker, M. Wolter, M. Schneider and A. Michaelis, J. Power Sources, 2019, 419, 119–126 CrossRef CAS .
135. H. Zheng, J. Li, X. Song, G. Liu and V. S. Battaglia, Electrochim. Acta, 2012, 71, 258–265 CrossRef CAS .
136. Y. Yang, R. Xu, K. Zhang, S. Lee, L. Mu, P. Liu, C. K. Waters, S. Spence, Z. Xu, C. Wei, D. J. Kautz, Q. Yuan, Y. Dong, Y. Yu, X. Xiao, H. Lee, P. Pianetta, P. Cloetens, J. Lee, K. Zhao, F. Lin and Y. Liu, Adv. Energy Mater., 2019, 9, 1900674 CrossRef .
137. Y.-H. Chen, C.-W. Wang, X. Zhang and A. M. Sastry, J. Power Sources, 2010, 195, 2851–2862 CrossRef CAS .
138. B. Vijayaraghavan, D. R. Ely, Y.-M. Chiang, R. García-García and R. E. García, J. Electrochem. Soc., 2012, 159, A548–A552 CrossRef CAS .
139. C.-J. Bae, C. K. Erdonmez, J. W. Halloran and Y.-M. Chiang, Adv. Mater., 2013, 25, 1254–1258 CrossRef CAS PubMed .
140. B. Delattre, R. Amin, J. Sander, J. De Coninck, A. P. Tomsia and Y.-M. Chiang, J. Electrochem. Soc., 2018, 165, A388–A395 CrossRef CAS .
141. L. Li, R. M. Erb, J. Wang, J. Wang and Y.-M. Chiang, Adv. Energy Mater., 2019, 9, 1802472 CrossRef .
142. I. V. Thorat, D. E. Stephenson, N. A. Zacharias, K. Zaghib, J. N. Harb and D. R. Wheeler, J. Power Sources, 2009, 188, 592–600 CrossRef CAS .
143. I. V. Thorat, D. E. Stephenson, N. A. Zacharias, K. Zaghib, J. N. Harb and D. R. Wheeler, J. Power Sources, 2009, 188, 592–600 CrossRef CAS .
144. J. Lim, Y. Li, D. H. Alsem, H. So, S. C. Lee, P. Bai, D. A. Cogswell, X. Liu, N. Jin and Y. Yu, Science, 2016, 353, 566–571 CrossRef CAS PubMed .
145. T. Nakamura, T. Watanabe, Y. Kimura, K. Amezawa, K. Nitta, H. Tanida, K. Ohara, Y. Uchimoto and Z. Ogumi, J. Phys. Chem. C, 2017, 121, 2118–2124 CrossRef CAS .
146. S. Xia, L. Mu, Z. Xu, J. Wang, C. Wei, L. Liu, P. Pianetta, K. Zhao, X. Yu, F. Lin and Y. Liu, Nano Energy, 2018, 53, 753–762 CrossRef CAS .
147. M. Fleckenstein, O. Bohlen, M. A. Roscher and B. Bäker, J. Power Sources, 2011, 196, 4769–4778 CrossRef CAS .
148. L. Mu, Q. Yuan, C. Tian, C. Wei, K. Zhang, J. Liu, P. Pianetta, M. M. Doeff, Y. Liu and F. Lin, Nat. Commun., 2018, 9, 2810 CrossRef PubMed .
149. L. Mu, R. Lin, R. Xu, L. Han, S. Xia, D. Sokaras, J. D. Steiner, T.-C. Weng, D. Nordlund, M. M. Doeff, Y. Liu, K. Zhao, H. L. Xin and F. Lin, Nano Lett., 2018, 18, 3241–3249 CrossRef CAS PubMed .

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ta06977a

‡ Equal contribution.

This journal is © The Royal Society of Chemistry 2019