Article Cite This: Acc. Chem. Res. 2018, 51, 89−96
pubs.acs.org/accounts
Electrochemical Characteristics of Layered Transition Metal Oxide Cathode Materials for Lithium Ion Batteries: Surface, Bulk Behavior, and Thermal Properties Published as part of the Accounts of Chemical Research special issue “Energy Storage: Complexities Among Materials and Interfaces at Multiple Length Scales”. Chixia Tian,†,‡ Feng Lin,† and Marca M. Doeff*,§ †
Department of Chemistry, Virginia Tech, Blacksburg, Virginia 24061, United States Macromolecules Innovation Institute (MII), Virginia Tech, Blacksburg, Virginia 24061, United States § Energy Storage and Distributed Resources Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States ‡
CONSPECTUS: Layered lithium transition metal oxides, in particular, NMCs (LiNixCoyMnzO2) represent a family of prominent lithium ion battery cathode materials with the potential to increase energy densities and lifetime, reduce costs, and improve safety for electric vehicles and grid storage. Our work has focused on various strategies to improve performance and to understand the limitations to these strategies, which include altering compositions, utilizing cation substitutions, and charging to higher than usual potentials in cells. Understanding the effects of these strategies on surface and bulk behavior and correlating structure−performance relationships advance our understanding of NMC materials. This also provides information relevant to the efficacy of various approaches toward ensuring reliable operation of these materials in batteries intended for demanding traction and grid storage applications. In this Account, we start by comparing NMCs to the isostructural LiCoO2 cathode, which is widely used in consumer batteries. Effects of changing the metal content (Ni, Mn, Co) upon structure and performance of NMCs are briefly discussed. Our early work on the effects of partial substitution of Al, Fe, and Ti for Co on the electrochemical and bulk structural properties is then covered. The original aim of this work was to reduce the Co content (and thus the raw materials cost) and to determine the effect of the substitutions on the electrochemical and bulk structural properties. More recently, we have turned to the application of synchrotron and advanced microscopy techniques to understand both bulk and surface characteristics of the NMCs. Via nanoscale-to-macroscale spectroscopy and atomically resolved imaging techniques, we were able to determine that the surfaces of NMC undergo heterogeneous reconstruction from a layered structure to rock salt under a variety of conditions. Interestingly, formation of rock salt also occurs under abuse conditions. The surface structural and chemical changes affect the charge distribution, the charge compensation mechanisms, and ultimately, the battery performance. Surface reconstruction, cathode/ electrolyte interface layer formation, and oxygen loss are intimately related, making it difficult to disentangle the effects of each of these phenomena. They are driven by the different redox activities of Ni and O on the surface and in the bulk; there is a greater tendency for charge compensation to occur on oxygen anions at particle surfaces rather than on Ni, whereas the Ni in the bulk is more redox active than on the surface. Finally, our latest research efforts are directed toward understanding the thermal properties of NMCs, which is highly relevant to their safety in operating cells.
■
INTRODUCTION
amount depending on the exact formulation, it is far less than that in LiCoO2. Most initial work focused on formulations in which the Ni and Mn contents were equal, in particular, LiNi1/3Mn1/3Co1/3O2 (often called NMC-333 or NMC-111).2 These compounds can be considered solid solutions of LiCoO2
LiCoO2 is commonly used in batteries powering small consumer devices such as cell phones and laptop computers, but the high cost and relative scarcity of cobalt1 prohibit its widespread use in larger devices intended for electric vehicles or grid storage. Layered transition metal oxides with the general formula LiNixCoyMnzO2 (where x + y + z ≈ 1), also known as NMCs, are technologically important cathode materials for lithium ion batteries. While NMCs still contain Co, with the © 2017 American Chemical Society
Received: October 18, 2017 Published: December 19, 2017 89
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research and LiNi0.5Mn0.5O2, the latter of which contains Ni in the divalent state and Mn in the tetravalent state.3 While the absence of Co in LiNi0.5Mn0.5O2 would, at first glance, seem to be advantageous, at least from a cost perspective, the material exhibits site disorder.4 The presence of about 10% Ni in Li sites and vice versa (ion-mixing)5 adversely affects diffusion of Li ions through the van der Waals gaps, resulting in rate limitations. Addition of Co reduces the ion mixing6 (in NMC-333, to about 2−3%),7 resulting in better rate performance. Hard X-ray absorption spectroscopy (XAS) experiments probing the transition metal K-edges of NMC-333 indicates that, like LiNi0.5Mn0.5O2, Ni is divalent and Mn tetravalent in the pristine states (Co is trivalent).8 In these compounds, Mn is generally redox-inactive, and Ni and Co are oxidized, ultimately, to the nominal tetravalent states upon cell charge. An increase in Ni and Mn content for these types of NMCs is advantageous in that the Co content is further reduced but results in increased ion mixing, which affects rate capabilities. While NMC-442 (LiNi0.4Mn0.4Co0.2O2)7 still exhibits good rate performance in spite of the increase in Ni content compared to NMC-333, formulations with lower Co contents (e.g., NMC-992, LiNi0.45Mn0.45Co0.1O2) exhibit rate limitations.9 Recently, a great deal of attention has been directed toward a class of NMCs known as Ni-rich NMCs, in which the Ni content exceeds that of the Mn.10,11 In these compounds, some Ni is present in the trivalent state in the pristine states, and when the Ni content is very high (e.g., NMC-811), the electrochemical behavior is reminiscent of LiNiO2, where all of the Ni is nominally trivalent, initially. These materials may also exhibit site disorder, in which Ni is located in Li sites. In general, the practical specific capacities increase with Ni content, rising to about 190−200 mAh/g for NMC-811, compared to 160 mAh/g for NMC-442, when using an upper cutoff voltage of 4.3 V in lithium half-cells. This makes the Nirich NMCs particularly interesting for vehicle applications, where energy density is of paramount importance. However, there are drawbacks in the form of shorter cycle life and poorer thermal stability, both of which need to be addressed before these materials can be successfully implemented for these applications. (Another category of NMCs known as LMRNMCs (for Li- and Mn-rich) are not discussed in this Account, although the interested reader can consult a recent publication for more information.12) Our interest in NMCs stems from their potential utility as cathodes in batteries for vehicle applications, where low cost and high energy density are of primary importance. To that end, our work was first directed toward decreasing Co content by partially substituting it with other metals, and understanding the effects on the electrochemical properties. A second goal was to improve practical utilization by cycling both substituted and unsubstituted NMCs to high potentials and understanding what limits the electrochemistry. Finally, our interests turned to understanding the thermal behavior of NMCs, which is particularly relevant to battery safety.
Figure 1. (top) Idealized representation of an NMC with the R3̅m structure. Yellow spheres represent Li ions, and transition metal cations are octahedrally coordinated by oxygen, shown in blue. (bottom) ADF/STEM image of NMC-442, showing the layered structure.
and method of synthesis,13 antisite or ion mixing, in which some Ni is located in Li (3a) sites and Li in 3b sites, can occur (a close inspection of the ADF/STEM image in Figure 1 reveals light spots, evidence of the presence of heavy atoms, in the darker layers where lithium should be located; lithium atoms are not detectable using this technique). Early computational work14 and some XAS studies15 of NMCs with symmetrical formulations suggested that, during charge, Ni is oxidized first, with participation by Co only above about 4.3 V vs Li/Li+. Because this is the usual upper voltage limit used in lithium half cells (corresponding to 4.2 V in full cells with graphite anodes), this suggested that it might be possible to partially or fully substitute low-cost metals such as Al or Fe for Co without adversely affecting practical capacities, provided that the structure is not substantially changed by the substitution. In fact, both partial Al and Fe substitution in NMC-33316 and NMC-44217 resulted in decreased capacities, with Fe substitution being particularly bad for performance, due to a greater degree of antisite mixing. Interestingly, although partial Al-substitution decreased practical capacities in both the 333 and 442 series, no matter what upper voltage limit cut off was used during charge, it appeared to improve rate capability. This could be attributed to a structural effect; the substitution results in an increase in the LiO2 slab spacing, which is helpful for lithium diffusion. Partial Al-substitution in the NMC-992 series also causes a monoclinic distortion for some compositions.18 While the NMC-992 compounds exhibited some rate limitations, probably due to ion-mixing, the Al substitution reduces changes in the lattice parameters during charge and discharge, which results in improved cycling for these cathodes compared to unsubstituted NMC-992. Finally, an in situ XAS study on a cell containing LiNi0.4Mn0.4Co0.15Al0.05O219 revealed that Co participates in redox processes early in the charge process as well as above 4.3 V. This may partially explain why the practical capacity decreased, although it is not clear whether the substitution itself resulted in the involvement of Co earlier in the delithiation compared to the parent compound.
■
STRUCTURE, ELECTROCHEMISTRY, AND EFFECTS OF SUBSTITUTION Figure 1 shows the idealized structure of an NMC compound, as well as an annular dark field/scanning transmission microscopy (ADF/STEM) image of an NMC-442 sample. The layered structure is clearly observable in the ADF/STEM image. Like LiCoO2, NMCs have the R3̅m structure (O3 stacking in the notation of Delmas). Depending on Ni content 90
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research The formation of solid solutions in stoichiometric NMCs where Ti replaces Co is limited to a narrow compositional range of a few percent.20 The substitution is aliovalent; Ti is tetravalent whereas Co is trivalent in the pristine materials, so that some kind of charge compensation is required for successful incorporation of Ti. Magnetic measurements suggest that Mn is partially reduced to the trivalent state in the Tisubstituted NMCs when the Li/TM ratio is roughly 1:1. Relatively little Mn3+ content can be tolerated in NMC structures, which may account for the limited solid solution range.21 The solid solution range appears to be somewhat greater for lithium-rich materials, because the charge compensation comes from inclusion of 2 Li ions per Ti ion substituted for Co, requiring no reduction of Mn to the trivalent state. Some Ti-substituted NMCs exhibit increased practical capacities compared to unsubstituted baseline compounds, particularly when charged to high potentials. This improvement is highly dependent upon method used to make the materials, type of precursors used, amount of Ti incorporated and other details of the synthesis but persists over many cycles. For example, a 3% Ti-substituted NMC-811 compound cycled to 4.5 V at 1C rate exhibited a capacity about 9% higher than the unsubstituted material initially.22 After 500 cycles, about 75% of the capacity was still retained, whereas the unsubstituted material had lost almost 40% of the original capacity and had exhibited much more severe voltage fade (consistent with increased cell impedance). Because Ti4+ is not expected to be electroactive in the normal operating voltage range where NMCs are cycled, the theoretical capacities of the Ti-substituted variants of NMCs are marginally less than that of the unsubstituted ones or roughly equal if Mn3+ redox activity can compensate for the Ti substitution. Thus, the observations of increased practical capacities are somewhat difficult to explain at first glance. The improvement appears to originate from better first cycle efficiencies (Figure 2), allowing higher discharge capacities to be obtained after charge,23 as well as a less steep rise in the charge voltage profile at moderate levels of delithiation.24 NMCs typically exhibit inefficiencies on the first cycle, no matter how conservative the charge limit is.25 (Subsequent cycles typically exhibit much better efficiencies, provided normal voltage limits are used). The loss can be recovered if discharges are driven to low potentials, indicating that the apparent irreversibility is not due to a bulk structural transformation, which would result in permanent loss. Instead, the inefficiency may be driven by changes in structure on particle surfaces and/or formation of interfacial layers, which impede diffusion of lithium especially during charge to high voltage. Typical behavior for half-cells containing Ti-substituted NMC cathodes charged to normal (4.3 V) and to high (4.7 V) potentials is shown in Figure 3.26 While capacities are higher initially for the cell charged to 4.7 V, the advantage is lost after just 20 cycles. The cycling to higher voltages is associated with increased cell impedance (Figure 3c), while the impedance and specific capacity of the cell cycled to 4.3 V remained stable. Nearly all of the capacity could be recovered for the cell charged to high voltages during subsequent cycling at C/50 rate, and a postmortem X-ray diffraction analysis of the NMC cathode failed to detect any bulk structural degradation or impurity phases. These observations taken together suggest strongly that surface processes are responsible for the rise in impedance and apparent capacity loss in the cells cycled to high potentials, rather than any bulk changes. Another clue is given
Figure 2. First cycles at 0.1 mA/cm2 between 4.7 and 2.0 V of lithium half cells containing NMC-333 (top), NMC-333 substituted with 2% Ti (middle), and Li-rich NMC-333 containing 3% Ti (bottom). Reproduced with permission from ref 23. Copyright 2011 Royal Society of Chemistry.
Figure 3. Cycling behavior of Li/NMC-442 cells (a) charged to 4.7 V at C/20 rate for 20 cycles (arrows designate gradual capacity fading with increasing cycle number) and (b) charged to 4.3 V at C/20 rate for 20 cycles. Discharge capacity as a function of cycle number for the two different cycling regimes is shown in panel c. Impedance data for pristine electrode, after 20 cycles to 4.3 and 4.7 V, and an electrode cycled 20 times to 4.7 V and washed with dimethyl carbonate, then reassembled into a cell are shown in panel d, and data for a pristine electrode exposed to electrolyte for 0 h, 5 h, 10 h, 1 day, 2 days, 3 days, 4 days, 5 days, and 7 days (arrow designates increasing time) are shown in panel e. Reproduced with permission from ref 26. Copyright 2014 Nature Publishing Group.
91
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research
provided by this technique, there is relatively little information about the exact form (e.g., crystal structure) that the reduced metal on the surface takes. ADF/STEM imaging, however, reveals that surface reconstruction to a rock salt (Fm3̅m) phase is responsible for the reduction in metal oxidation states (Figure 5). The great virtue of the microscopy is the ability to
in Figure 3e, showing data for an electrode aged in electrolytic solution but not subjected to any electrochemical cycling. A gradual rise in impedance over time is observed. The surface and interfacial processes responsible for these phenomena are discussed in the next section.
■
SURFACE AND INTERFACIAL CHARACTERISTICS In order to understand the surface chemistry, soft XAS experiments were performed on NMC electrodes at different states-of-charge, harvested from lithium half-cells.27 These experiments probe the electronic structures of the transition metals to various sample depths, depending on detection mode (Figure 4). At the Stanford Synchrotron Radiation Lightsource,
Figure 5. ADF/STEM image of an NMC particle after one cycle in an electrochemical cell to 4.7 V. The rock salt structure on surfaces is clearly evident, with the thickness showing facet-dependence. Adapted with permission from ref 26. Copyright 2014 Nature Publishing Group.
reveal much-needed detail complementary to the soft XAS experiment, but a weakness can be that the particle under observation may not be typical of the sample as a whole. In this case, a number of different particles were examined, revealing that the surface reconstruction was very heterogeneous. Not only did some particles exhibit much more surface reconstruction than others, but the thicknesses of the reconstruction layers varied with the facet, tending to be thicker on those approximately normal to ⟨001⟩ or the direction of lithium diffusion (this can be readily observed in Figure 5). Surface reconstruction occurs under a variety of conditions including normal cycling, charging to high potentials, and even just exposure to the electrolytic solution for moderate amounts of time (tens of hours) or aging in the absence of electrolytic solution, without electrochemical cycling. The degree to which it occurs is also a function of how the material is synthesized.28 The amount of reduced metal on particle surfaces appears to increase with cycle number, at least up to about 20 cycles, with more severe effects observed when high voltage limits are used compared to 4.3 V. Interestingly, Ti-substitution appears to mitigate these effects.29 Density Functional Theory (DFT) calculations predict that conversion to a rock salt structure from the R3̅m structure of NMCs should occur spontaneously beyond a certain threshold of delithiation. Figure 6 shows free energies for NMC-333 and Ti-substituted NMC-333 as a function of state-of-charge. Rock salt formation is thermodynamically predicted to occur at or below a free energy of 0, as indicated by the dashed line; in the case of NMC-333, at about 70% removal of lithium.24 The energy of formation of rock salt is increased for the Tisubstituted variant because Ti more strongly binds oxygen compared to Ni, Mn, or Co, which stabilizes the delithiated NMC structure.
Figure 4. (a) Schematic of different detection modes for soft XAS experiments: Auger electron yield (AEY), total electron yield (TEY), and fluorescence yield (FY). (b) Ni L-edge spectra in FY mode of NMC electrodes at different states-of-charge. (c) Ni L-edge spectra of an electrode charged to 4.7 V and (d) L3high/L3low ratios taken from panel c as a function of probing depth. Reproduced with permission from ref 27. Copyright 2014 the Royal Society of Chemistry.
where these experiments were performed, Auger electron yield (AEY) mode probes the very top surface of the electrode, about 1−2 nm in, while total electron yield (TEY) samples about 2−5 nm in, and fluorescence mode (FY) about 50−100 nm in, for example, the bulk. Figure 4b shows the Ni L-edge data in FY mode of NMC electrodes at different states-of-charge. The ratios of the L3high/L3low peaks directly correlate to the average Ni oxidation state (although only relative, not absolute values can be obtained). Ni is first oxidized and then reduced as the electrode is charged and discharged, as reflected in the changes in the ratios of the L3 peaks seen in Figure 4b. Figure 4c shows Ni L-edge data in the three different modes for an electrode charged to 4.7 V in a lithium half-cell. Surprisingly, the Ni oxidation state at the very surface of the electrode (AEY mode) is lower than that 5 nm in, and in the bulk (Figure 4d). Further experiments probing the Mn and Co L-edges reveal that these metals are also reduced on particle surfaces under these same conditions. The soft XAS experiment probes many particles in the electrode because of the relatively large spot size used (∼1 mm2); in other words, it is ensemble-averaged. Although details of the electronic structure as a function of sample depth are 92
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research
In actuality, while the structural conversion does appear to increase with number of cycles to high potentials, it is generally limited to particle surfaces rather than extending into the bulk. The transformation to rock salt requires loss of oxygen, and is probably kinetically limited by the sluggishness of oxygen diffusion through the bulk of the NMC lattice. Interestingly, the degree of surface reconstruction observed upon cycling is also a function of surface Ni content. NMC-442 synthesized by a spray pyrolysis method appears to retain capacity better upon high voltage cycling than samples made by co-precipitation.30 A combined synchrotron Transmission Xray Microscopy (TXM)/STEM-EELS study revealed that, while the global compositions of the two materials were identical, the spray pyrolyzed sample had a heterogeneous distribution of metals throughout the particles, while the coprecipitated one was more homogeneous. Specifically, the Ni content on spray pyrolyzed particle surfaces was less than that in the bulk, not only for the large secondary particles (agglomerates) but also for the nanometric primary particles, which make up the agglomerates (Figure 7). The STEM-EELS data in Figure 7d show that there is decreased Ni content at the surface of the primary particles compared to the bulk. The lower surface Ni content correlates to a lesser amount of reduced metal on particle surfaces after 20 cycles to 4.7 V and improved cycling (Figure 8) compared to the co-precipitated NMC.
Figure 6. Calculated free energy of formation at room temperature for the rock salt structure from NMC-333 (black line, Ti00) and 3% Tisubstituted NMC-333 (red line, Ti03). Reproduced with permission from ref 24. Copyright 2014 American Chemical Society.
In principle, this indicates that repeated cycling to 4.7 V vs Li+/Li (or storage of electrodes charged to this potential) should result in complete transformation to rock salt eventually.
Figure 7. (a) SEM image showing primary particles of NMC-442 prepared by spray pyrolysis. The white box encloses a particle−particle interface detailed in the ADF-STEM image shown in panel b. The rainbow arrow shows the scan direction for Mn, Co, and Ni L-edge EELS data shown in panel c. In panel d, Ni L3 integrated areas as a function of scan position are shown. Adapted with permission from ref 30. Copyright 2016 Nature Publishing Group. 93
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research
associated with capacity fading during cycling, or if other factors, such as formation of resistive cathode/electrolyte interphases (CEIs) are to blame instead.33 It is difficult to disentangle effects of CEI formation, surface reconstruction, and oxygen loss, since these phenomena are related to one another. Under normal cycling conditions where capacity fading is minimal, surface reconstruction still occurs but progresses much more slowly than during high voltage cycling, where losses are more clearly evident.26 Similarly, the formation of surface reaction layers or CEIs are more apparent during high voltage cycling than during normal cycling or storage in electrolytic solutions (Figure 9).26,27 STEM-EELS26 and synchrotron XPS27 analysis of these reaction layers reveal the presence of both inorganic (e.g., LiF) and organic components.
Figure 8. (a) Mn L-edge soft XAS spectra for pristine NMC-442 (bottom, in blue), spray pyrolyzed NMC-442 (middle, red), and coprecipitated NMC-442 (top, green) cycled 20 times to 4.7 V vs Li+/Li and stopped in the discharged state and (b) Co L-edge soft XAS spectra for the same three samples. Reproduced with permission from ref 30. Copyright 2016 Nature Publishing Group.
Figure 9. (a) STEM image of an NMC particle stored in electrolytic solution for 7 days and (b) STEM image of an NMC particle cycled 20 times to 4.7 V vs Li+/Li. Reproduced with permission from ref 26. Copyright 2014 Nature Publishing Group.
A recent quantitative analysis of O K-edge and transition metal L-edge soft XAS data obtained in both TEY and FY modes on partially and fully charged and discharged NMC-622 electrodes revealed a depth-dependent charge compensation mechanism.31 The transition metal L-edge data showed that both Ni and Co participate in redox processes for this Ni-rich NMC. However, a comparison of TEY and FY data showed that Ni was consistently more oxidized in the bulk compared to that on the surface, particularly at high states-of-charge. This discrepancy was not observed for Co. The O K-edge data reveals information about the TM 3d−O 2p hybridization, with the integrated intensity of peaks in the relevant region proportional to the number of hole states in the oxygen, reflecting O redox processes. Here, the opposite was true, with charge compensation via O redox activity higher on surfaces compared to the bulk. The higher reactivity of surface oxygen when Ni is present explains the strong tendency for surface reconstruction with concomitant loss of oxygen in NMC materials, with the degree roughly correlated to the amount of Ni on particle surfaces. It is significant that NMC-622 subjected to partial chemical delithiation with NO2BF4 exhibits much less surface reconstruction compared to NMC-622 samples that were electrochemically delithiated and relithiated to roughly the same degree, as evidenced by soft XAS analysis.32 This indicates that the electrolytic solution must play a role in this phenomenon, probably by acting as a source of electrons for the reduction of metals on particle surfaces. In contrast, no obvious reducing agent is available under the conditions of the chemical oxidation. The question has been raised recently whether surface reconstruction is responsible for the rise of impedance
There is a strong tendency for NMCs to form rock salt phases under conditions of abuse as well. For example, this occurs when cycled materials are subjected to electron beams at either low scan rates or high dosages.34 Caution should be used to limit electron doses when using transmission electron microscopy to examine NMC samples to prevent beam damage. It is perhaps not a coincidence, however, that beam damage induces the same sort of phase transitions as those that occur during other types of abuse, such as high voltage cycling or thermal excursions, as discussed in the next section.
■
THERMAL PROPERTIES The thermal stability of NMCs is a function of Ni-content35 as well as state-of-charge. A recent in situ XRD study36 on NMC622 delithiated to different degrees shows that the thermal stability decreases as the lithium content is lowered. As lithium content is decreased or the temperature is increased, progressively more reduced metal oxides are produced (first LiM2O4 type spinels, then M3O4 type spinels, and finally MOtype rock salts). This progression is accompanied by an increasing degree of oxygen loss, which has grave safety implications for the operation of cells. Strategies to ensure thermal abuse tolerance are clearly needed and are dependent upon a thorough understanding of how these reactions evolve. To this end, we are currently undertaking a comprehensive study of the bulk and surface properties of NMCs as a function of temperature and state-of-charge. In addition to XRD studies of the crystal structure changes of delithiated NMCs during heating, we further investigated the redox phase transformation of the materials by using an integrative approach to map the valence states of thermally 94
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research treated delithiated NMCs.37 Construction of 3D topographies for the valence states of transition metal cations reveal that the phase propagation in the NMC particles follows a complex pathway that is defined by the local curvature of valence state topographies. This can be related to the grain boundaries and/ or compositional variation within the local chemical environments. Our studies imply not only that the lithium amount (i.e., state of charge) and NMC particle composition play vital role in phase change under thermal treatment but also that the local chemical environment on the mesoscale influences the thermal stability.
expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof or the Regents of the University of California. The authors declare no competing financial interest. Biographies Chixia Tian was born in Yichang, Hubei, China. She received her Bachelor’s degree in Materials Science and Engineering with a specialty of Polymer from Beijing University of Chemical Technology, China, in 2009. Then she received her Ph.D. in Chemistry from Colorado school of Mines in 2015. After two years of postdoctoral training with Dr. Marca Doeff at Lawrence Berkeley National Laboratory, she is currently working at Virginia Tech as a research scientist in the Chemistry department. Her research interests encompass nanomedicine, cathode materials, and polymer electrolyte materials for lithium batteries.
■
SUMMARY The electrochemical properties of NMC cathode materials are a function of composition (Ni, Co, and Mn content as well as substituents such as Al, Fe, or Ti), details of their synthesis, and the voltage limits used to charge the cells. Compositional variations affect the degree of ion mixing, which, in turn, determine practical capacities and rate capabilities. Substitutions affect lattice parameter changes encountered during cycling, influencing the capacity retention. The tendency for surface reconstruction to a rock salt phase to occur is also a function of surface Ni content, as well as the charge voltage limit. While first principle calculations predict that layered NMCs should spontaneously convert to rock salt upon delithiation, in practice this happens only on the surface rather than the bulk, due to kinetic limitations. Surface reconstruction, however, is probably responsible for the first cycle inefficiencies commonly seen in NMC cells, no matter what voltage limit is used but is decreased when Ti-substituted materials are used. Cycling to high potentials results in increased surface reconstruction with concomitant oxygen loss and formation of a resistive cathode/electrolyte interface layer. This results in impedance rise, which is responsible for the capacity fading observed during high voltage cycling. The tendency to form rock salt is also observed under abuse conditions such as beam damage from TEM measurements and thermal treatment of delithiated samples.
■
Feng Lin was born in Fuding, Fujian, China. He is currently an Assistant Professor of Chemistry, with a courtesy appointment in the Department of Materials Science and Engineering, at Virginia Tech. He holds Bachelor’s degree in Materials Science and Engineering (2009) from Tianjin University and a M.Sc. degree (2011) and Ph.D. degree (2012) in Materials Science from Colorado School of Mines. Prior to Virginia Tech, he worked at QuantumScape Corporation as a Senior Member of the Technical Staff, Lawrence Berkeley National Lab as a Postdoctoral Fellow, and National Renewable Energy Lab as a Graduate Research Assistant. He is fascinated by energy sciences, materials electrochemistry, and synchrotron X-rays. Marca M. Doeff was born in Leiden, The Netherlands, and received her B.A. in Chemistry from Swarthmore College, Swarthmore, PA, in 1978 and a Ph.D. in Inorganic Chemistry from Brown University, Providence, RI, in 1983. She has held the position of Staff Scientist at Lawrence Berkeley National Laboratory, Berkeley, CA, since 1990. Her research interests focus on materials for energy storage, including electrode materials for sodium and lithium ion batteries and solid electrolytes for lithium batteries. She is currently a fellow of the Royal Society of Chemistry and of the Electrochemical Society.
■
ACKNOWLEDGMENTS This work was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Office of Vehicle Technologies of the U.S. Department of Energy, under Contract No. DE-AC02-05CH11231. F.L. gratefully acknowledges Virginia Tech Department of Chemistry startup funds and ORAU Ralph E. Powe Junior Faculty Enhancement Award. C.T. acknowledges the support from Virginia Tech College of Science. The authors also acknowledge Gregory Steward of SLAC Communication for assistance with the artwork in the Conspectus.
AUTHOR INFORMATION
Corresponding Author
*E-mail: mmdoeff@lbl.gov. ORCID
Feng Lin: 0000-0002-3729-3148 Marca M. Doeff: 0000-0002-2148-8047 Notes
This document was prepared as an account of work sponsored by the United States Government. While this document is believed to contain correct information, neither the United States Government nor any agency thereof, nor the Regents of the University of California, nor any of their employees, makes any warranty, express or implied, or assumes any legal responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by its trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof, or the Regents of the University of California. The views and opinions of authors
■
REFERENCES
(1) Nelson, E. The hottest thing in the markets right now is an obscure metal mined in DR Congo. https://qz.com/969115/cobalt-isthe-unlikely-hero-of-financial-markets-right-now/ (accessed Nov. 30, 2017). (2) Yabuuchi, N.; Ohzuku, T. Novel Lithium Insertion Material of LiCo1/3Ni1/3Mn1/3O2 for Advanced Lithium-Ion Batteries. J. Power Sources 2003, 119−121, 171−174. (3) Yoon, W.-S.; Balasubramanian, M.; Yang, X.-Q.; Fu, Z.; Fischer, D. A.; McBreen, J. Soft X-ray Absorption Spectroscopic Study of a LiNi0.5Mn0.5O2 Cathode During Charge. J. Electrochem. Soc. 2004, 151, A246−A251. (4) Van der Ven, A.; Ceder, G. Ordering in Lix(Ni0.5Mn0.5)O2 and its Relation to Charge Capacity and Electrochemical Behavior in 95
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
Article
Accounts of Chemical Research Rechargeable Lithium Batteries. Electrochem. Commun. 2004, 6, 1045− 1050. (5) Breger, J.; Dupre, N.; Chupas, P. J.; Lee, P. L.; Proffen, T.; Parise, J. B.; Grey, C. P. Short and Long-Range Order in the Positive Electrode Material, Li(NiMn)0.5O2: A Joint X-ray and Neutron Diffraction Pair Distribution Function Analysis and NMR Study. J. Am. Chem. Soc. 2005, 127, 7529−7537. (6) Chernova, N. A.; Ma, M.; Xiao, J.; Whittingham, M. S.; Breger, J.; Grey, C. P. Layered LixNiyMnyCo1−2yO2 Cathodes for Lithium Ion Batteries: Understanding Local Structure via Magnetic Properties. Chem. Mater. 2007, 19, 4682−4693. (7) Li, Z.; Chernova, N. A.; Roppolo, M.; Upreti, S.; Petersburg, C.; Alamgir, F. M.; Whittingham, M. S. Comparative Study of the Capacity and Rate Capability of LiNiyMnyCo1−2yO2 (y = 0.5, 0.45, 0.4, 0.33). J. Electrochem. Soc. 2011, 158, A516−A522. (8) Yoon, W.-S.; Balasubramanian, M.; Chung, K. Y.; Yang, X.-Q.; McBreen, J.; Grey, C. P.; Fischer, D. A. Investigation of the Charge Compensation Mechanism on the Electrochemically Li-Ion Deintercalated Li1‑xCo1/3Ni1/3Mn1/3O2 Electrode System by Combination of Soft and Hard X-ray Absorption Spectroscopy. J. Am. Chem. Soc. 2005, 127, 17479−17487. (9) Conry, T. E.; Mehta, A.; Cabana, J.; Doeff, M. M. XAFS Investigation of LiNi0.45Mn0.45Co0.1‑yAlyO2 Positive Electrode Materials. J. Electrochem. Soc. 2012, 159, A1562−A1571. (10) Liu, W.; Oh, P.; Liu, X.; Lee, M.-J.; Cho, W.; Chae, S.; Kim, Y.; Cho, J. Nickel-Rich Layered Lithium Transitional-Metal Oxide for High-Energy Lithium-Ion Batteries. Angew. Chem., Int. Ed. 2015, 54, 4440−4457. (11) Xu, J.; Lin, F.; Doeff, M. M.; Tong, W. A Review of Ni-based Layered Oxides for Rechargeable Li-Ion Batteries. J. Mater. Chem. A 2017, 5, 874−901. (12) Croy, J. R.; Balasubramanian, M.; Gallagher, K. G.; Burrell, A. K. Review of the Department of Energy’s “Deep Dive” Effort to Understand Voltage Fade in Li- and Mn-rich Cathodes. Acc. Chem. Res. 2015, 48, 2813−2821. (13) Patoux, S.; Doeff, M. M. Direct Synthesis of LiNi1/3Co1/3Mn1/3O2 from Nitrate Precursors. Electrochem. Commun. 2004, 6, 767−772. (14) Hwang, B. J.; Tsai, Y. W.; Carlier, D.; Ceder, G. A Combined Computational/Experimental Study on LiNi1/3Co1/3Mn1/3O2. Chem. Mater. 2003, 15, 3676−3682. (15) Tsai, Y. W.; Hwang, B. J.; Ceder, G.; Sheu, H. S.; Liu, D. G.; Lee, J. F. In-Situ X-ray Absorption Spectroscopic Study on Variation of Electronic Transitions and Local Structure of LiNi1/3Co1/3Mn1/3O2 Cathode Material during Electrochemical Cycling. Chem. Mater. 2005, 17, 3191−3199. (16) Wilcox, J.; Patoux, S.; Doeff, M. Structure and Electrochemistry of LiNi1/3Co1/3Mn1/3O2 (M = Ti, Al, Fe) Positive Electrode Materials. J. Electrochem. Soc. 2009, 156, A192−A198. (17) Wilcox, J. D.; Rodriguez, E. E.; Doeff, M. M. The Impact of Aluminum and Iron Substitution on the Structure and Electrochemistry of Li(Ni0.4Co0.2‑yMn0.4)O2 Materials. J. Electrochem. Soc. 2009, 156, A1011−A1018. (18) Conry, T. E.; Mehta, A.; Cabana, J.; Doeff, M. M. Structural Underpinnings of the Enhanced Cycling Stability upon Al-substitution in LiNi0.45Mn0.45Co0.1‑yAlyO2 Positive Electrode Materials for Li-ion Batteries. Chem. Mater. 2012, 24, 3307−3317. (19) Rumble, C.; Conry, T. E.; Doeff, M.; Cairns, E. J.; Penner-Hahn, J. E.; Deb, A. Structural and Electrochemical Investigation of Li(Ni0.4Co0.15Al0.05Mn0.4)O2 Cathode Material. J. Electrochem. Soc. 2010, 157, A1317−A1322. (20) Kam, K. C.; Mehta, A.; Heron, J. T.; Doeff, M. M. Electrochemical and Physical Properties of Ti-Substituted Layered Nickel Manganese Cobalt Oxide (NMC) Cathode Materials. J. Electrochem. Soc. 2012, 159, A1383−A1392. (21) Xiao, J.; Chernova, N. A.; Whittingham, M. S. Influence of Manganese Content on the Performance of LiNi0.9‑yMnyCo0.1O2 (0.45 ≤ y ≤ 0.6) as a Cathode Material for Li-Ion Batteries. Chem. Mater. 2010, 22, 1180−1185.
(22) Steiner, J.; Lin, F. Unpublished results. (23) Kam, K. C.; Doeff, M. M. Aliovalent Titanium Substitution in Layered Mixed Li Ni-Mn-Co Oxides for Lithium Battery Applications. J. Mater. Chem. 2011, 21, 9991−9993. (24) Markus, I. M.; Lin, F.; Kam, K. C.; Asta, M.; Doeff, M. M. Computational and Experimental Investigation of Ti Substitution in Li1(NixMnxCo1−2x‑yTiy)O2 for Lithium Ion Batteries. J. Phys. Chem. Lett. 2014, 5, 3649−3655. (25) Kang, S.-H.; Yoon, W.-S.; Nam, K.-W.; Yang, X.-Q.; Abraham, D. P. Investigating the First-Cycle Irreversibility of Lithium Metal Oxide Cathodes for Li Batteries. J. Mater. Sci. 2008, 43, 4701−4706. (26) Lin, F.; Markus, I. M.; Nordlund, D.; Weng, T.-C.; Asta, M. D.; Xin, H. L.; Doeff, M. M. Surface Reconstruction and Chemical Evolution of Stoichiometric Layered Cathode Materials for LithiumIon Batteries. Nat. Commun. 2014, 5, 3529. (27) Lin, F.; Nordlund, D.; Markus, I. M.; Weng, T.-C.; Xin, H. L.; Doeff, M. M. Profiling the Nanoscale Gradient in Stoichiometric Layered Cathode Particles for Lithium-Ion Batteries. Energy Environ. Sci. 2014, 7, 3077−3085. (28) Lin, F.; Nordlund, D.; Pan, T.; Markus, I. M.; Weng, T.-C.; Xin, H. L.; Doeff, M. M. Influence of Synthesis Conditions on the Surface Passivation and Electrochemical Behavior of Layered Cathode Materials. J. Mater. Chem. A 2014, 2, 19833−19840. (29) Wolff-Goodrich, S.; Lin, F.; Markus, I. M.; Nordlund, D.; Xin, H. L.; Asta, M.; Doeff, M. M. Tailoring the Surface Properties of LiNi0.4Mn0.4Co0.2O2 by Titanium Substitution for Improved High Voltage Cycling Performance. Phys. Chem. Chem. Phys. 2015, 17, 21778−21781. (30) Lin, F.; Nordlund, D.; Li, Y.; Quan, M. K.; Cheng, L.; Weng, T.C.; Liu, Y.; Xin, H. L.; Doeff, M. M. Metal Segregation in Hierarchically Structured Cathode Materials for High-Energy Lithium Batteries. Nature Energy 2016, 1, 15004. (31) Tian, C.; Nordlund, D.; Xin, H.; Xu, Y.; Liu, Y.; Sokara, D.; Lin, F.; Doeff, M. M. Depth-Dependent Chemical Environment Drives the Charge Compensation and Transformation of Nickel-Rich Layered Cathode Materials. J. Phys. Chem. C 2017, submitted for publication. (32) Tian, C.; Xu, Y.; Nordlund, D.; Sun, Z.; Liu, Y.; Doeff, M. M. Charge Heterogeneity and Surface Chemistry in Polycrystalline Cathode Materials. Joule 2017, in press. (33) Li, J.; Liu, H.; Xia, J.; Cameron, A. R.; Nie, M.; Botton, G. A.; Dahn, J. R. The Impact of Electrolyte Additives and Upper Cut-off Voltage on the Formation of a Rocksalt Surface Layer in LiNi0.8Mn0.1Co0.1O2 Electrodes. J. Electrochem. Soc. 2017, 164, A655−A665. (34) Lin, F.; Markus, I. M.; Doeff, M. M.; Xin, H. L. Chemical and Structural Stability of Lithium-Ion Battery Electrode Materials under Electron Beam. Sci. Rep. 2013, 4, 5694. (35) Noh, H.-J.; Youn, S.; Yoon, C. S.; Sun, Y.-K. Comparison of the Structural and Electrochemical Properties of Layered Li[NixCoyMnz]O2 (x = 1/3, 0.5, 0.6, 0.7, 0.8 and 0.85) Cathode Material for LithiumIon Batteries. J. Power Sources 2013, 233, 121−130. (36) Tian, C.; Nordlund, D.; Liu, Y.; Sokaras, D.; Doeff, M. M. Unpublished results. (37) Mu, L.; Yuan, Q.; Tian, C.; Pianetta, P.; Doeff, M. M.; Liu, Y.; Lin, F. Propagation Topography of Redox Phase Transformations in Heterogeneous Layered Oxide Cathode Materials. Nature Commun. 2017, submitted for publication.
96
DOI: 10.1021/acs.accounts.7b00520 Acc. Chem. Res. 2018, 51, 89−96
