Probing the Degradation Mechanism of Li2MnO3 Cathode for Li-Ion

Jan 14, 2015 - ... of Electronic Science and Technology of China, Chengdu, 610054, P.R. China. ⊥ Department of Mechanical Engineering and Materials ...
0 downloads 11 Views 2MB Size
Download PDF
Article pubs.acs.org/cm

Probing the Degradation Mechanism of Li2MnO3 Cathode for Li-Ion Batteries Pengfei Yan,† Liang Xiao,‡,§ Jianming Zheng,‡ Yungang Zhou,∥ Yang He,⊥ Xiaotao Zu,∥ Scott X. Mao,⊥ Jie Xiao,‡ Fei Gao,# Ji-Guang Zhang,*,‡ and Chong-Min Wang*,† †

Environmental Molecular Sciences Laboratory and ‡Energy and Environmental Directorate, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, Washington 99352, United States § Department of Chemistry, School of Chemistry, Chemical Engineering and Life Sciences, Wuhan University of Technology, Wuhan, Hubei 430070, China ∥ Department of Applied Physics, University of Electronic Science and Technology of China, Chengdu, 610054, P.R. China ⊥ Department of Mechanical Engineering and Materials Science, University of Pittsburgh, 3700 O’Hara Street, Pittsburgh, Pennsylvania 15261, United States # Department of Nuclear Engineering and Radiological Sciences, University of Michigan, Ann Arbor, Michigan 48109, United States S Supporting Information *

ABSTRACT: Capacity and voltage fading of Li2MnO3 is a major challenge for the application of this category of material, which is believed to be associated with the structural and chemical evolution of the materials. This paper reports the detailed structural and chemical evolutions of Li2MnO3 cathode captured by using aberration corrected scanning/transmission electron microscopy (S/TEM) after certain numbers of charge−discharge cycling of the batteries. It is found that structural degradation occurs from the very first cycle and is spatially initiated from the surface of the particle and propagates toward the inner bulk as the cyclic number increases, featuring the formation of the surface phase transformation layer and gradual thickening of this layer. The structure degradation is found to follow a sequential phase transformation: monoclinic C2/m → tetragonal I41 → cubic spinel, which is consistently supported by the decreasing lattice formation energy based on DFT calculations. For the first time, high spatial resolution quantitative chemical analysis reveals that 20% oxygen in the surface phase transformation layer is removed and such a newly developed surface layer is a Li-depleted layer with reduced Mn cations. This work demonstrates a direct correlation between structural degradation and the cell’s electrochemical degradation, which enhances our understanding of Li−Mn-rich (LMR) cathode materials.

■

INTRODUCTION Lithium ion battery (LIB) is a key resource for mobile energy and one of the most promising solutions for environmentfriendly transportation.1,2 However, there are still some critical issues that need to be addressed in order to satisfy commercial application requirements, such as typically energy density, rate capability, cycling performance, and safety problems.3−6 For current lithium ion batteries, the cathode materials are the limiting factor, which determines the working voltage and capacities. Recently, layered lithium manganese rich (LMR) cathode materials have attracted much attention due to their high capacities and high working voltage as compared with traditional spinel LiMn2O4 and LiFePO4 cathode materials.6−13 Generally, LMR cathode materials have the expression xLiMO2·(1 − x)[Li2MnO3] (M = Mn, Ni, Co, 0 ≤ x < 1). As the parent component of LMR, Li2MnO3 cathode materials also attracted much attention for many reasons. First, Li2MnO3 itself has a very high theoretical capacity. Thus, many efforts © 2015 American Chemical Society

have been done in order to activate and utilize all the Li-ions in Li2MnO3 material.14−17 Second, it can enhance our fundamental understanding of the electrochemistry of Mn4+containing cathode materials.6,13,18−20 Third, we can obtain the knowledge necessary for designing new LMR cathode materials.10,21 The Li2MnO3 cathode was initially believed to be electrochemically inert because all the Mn cations are Mn4+ in this material until Kalyani et al. demonstrated that Li2MnO3 could be electrochemically activated in 1999.22 The widely accepted mechanisms to explain the delithiation process include simultaneous removal of Li and O 23,24 and Li + −H + exchange,25,26 but which one plays the dominant role is still under debate. It has been claimed that prolonged cycling could Received: November 19, 2014 Revised: January 13, 2015 Published: January 14, 2015 975

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

Figure 1. (a) Left side shows charge and discharge capacities as a function of cycle numbers at 2.0−4.8 V. Right side shows Coulombic efficiency for each cycle. (b) Charge−discharge profiles at 1st, 10th, and 45th cycles as indicated by arrows in (a).

transform Li2MnO3 to LiMn2O4-spinel.25,27 However, recently X-ray absorption spectroscopy results from the cycled sample are not consistent with the LiMn2O4-spinel structure; instead, the P3 structure was proposed.28 In a chemically delithiated Li2MnO3 sample, Wang et al. observed the formation of the MnO2 phase.29 Up to now, it is still not clear which structure is formed after charge−discharge cycles and how the structure evolves during the cycling of Li2MnO3 cathode. In this work, we present the results of systematic observations on charge−discharge cycled Li2MnO3 samples with different cycle numbers using aberration corrected scanning/transmission electron microscope (S/TEM) and DFT calculations. By comparing the 1-cycle, 10-cycles, and 45-cycles samples with the pristine sample, we confirmed structure transformation was initiated from the particle surface and propagated into the inner bulk upon continuous cycling, where the cycling induced structure transformation sequence is monoclinic C2/m → tetragonal I41 → cubic spinel. The Li is gradually depleted in the transformed layer, and for the first time we confirmed 20% oxygen has been removed in the transformation layer. Such structure degradation is believed to be the main reason for fast degradation of the batteries.

■

for TEM microanalysis. The pristine and cycled samples were studied using a probe aberration-corrected S/TEM microscope (FEI Titan 80300 operated at 300 kV, JEOL JEM-ARM200CF operated at 200 kV). Selective area electron diffraction (SAED), high resolution transmission electron microscopy (HRTEM), STEM high angle annular dark field (HAADF) observations, and electron energy loss spectroscopy (EELS) were conducted on an FEI Titan 80-300 microscope to probe the structure and chemical evolution at multiple scales. The STEM-HAADF images were collected by the annular detector in the range of 55−220 mrad, and the EELS were acquired using Gatan Image Filter (GIF, Quantum 965) with collection semiangle ∼ 50 mrad. The electron beam has a convergence angle of 17.8 mrad. The EELS data were collected from a thin area to reduce multiple inelastic scatterings. A low-loss spectrum and the core-loss spectrum were collected from the same area at the same time. The low-loss spectrum was used to remove the multiple inelastic scattering effect in the coreloss region using the Fourier ratio technique (DigitalMicrograph, Gatan Inc.). Nano beam diffraction was performed at 300 kV microprobe scan mode with a 10 μm C2 aperture. The probe diameter was measured to be less than 3 nm (full width at half-maximum) with ∼1.1 mrad convergence angle. Energy dispersive X-ray spectroscopy (EDS) was conducted on a JEOL JEM-ARM200CF microscope equipped with a JEOL SDD-detector with a 100 mm2 X-ray sensor, enabling 10 times faster X-ray collections than a traditional detector with excellent noise-to-signal ratio. The EDS quantification was performed using Analysis Station 3.8.0.52 (JEOL Engineering Co., Ltd.).

EXPERIMENTAL SECTION

■

Material Synthesis. The Li2MnO3 samples were prepared by ball milling (SPEX SamplerPrep 8000 M Mixer/Mill) the mixture of Li2CO3 and MnO2 (all from Sigma-Aldrich) in the stoichiometric amount for 4 h followed by a heat treatment at 900 °C for 12 h in air. To compensate for Li loss at high temperature, 5 wt % extra amount of Li2CO3 was used in the starting materials. Electrochemical Test. Electrodes were prepared by casting a slurry of the Li2MnO3 samples, Super P (from Timcal), and poly(vinylidene fluoride) (pVDF, Kynar HSV900, Arkema Inc.) in an N-methyl pyrrolidone (Sigmal-Aldrich) solvent onto aluminum foil. The weight ratio of Li2MnO3:Super P:pVDF was 80:10:10. After drying at 80 °C, the electrodes were punched into disks in the diameter of 1.4 cm. The active material loading was 3−4 mg cm−2. 2032 coin cells with aluminum clad cans were assembled in an argonfilled glovebox (MBraun) using lithium foil as anode, Celgard polyethylene film as separator, and 1 M LiPF6 in a mixture of ethylene methyl carbonate and ethylene carbonate at a 7:3 volume ratio as electrolyte. The electrochemical tests were performed on an Arbin BT-2000 battery tester at room temperature. The cells were cycled between 2.0 and 4.8 V vs Li/Li+ at 10 mA g−1. Material Characterization. After cycling, the obtained electrode was washed by ethylene carbonate three times and dried in vacuum for 12 h. The electrodes were peeled off from the Al foil and ground to fine powders. The powder particles were dusted on a lacey carbon grid

RESULTS AND DISCUSSION Electrochemical Performance. The electrochemical performance of the coin cell made with Li2MnO3 as cathode material is shown in Figure 1. Typically, the capacity of the cell increases initially with the progression of the cycling and reaches the maximum at 9 cycles, followed by a continuously decreasing with further cycling. Due to cell’s fast degradation, its Coulombic efficiency is very low as plotted in Figure 1a. Figure 1b shows the voltage profile at the 1st, 10th, and 45th cycles. The electrochemical performance of the Li2MnO3 coin cell as illustrated in Figure 1 indicates that this cell needs several cycles to be fully activated, and as soon as it is fully activated, it suffers from a quick fading in both capacity and working voltage, which are closely related to the structural and chemical evolution of the materials as described in the following section. Materials Characterization. Comparing with the pristine Li 2 MnO 3 cathode, we clearly observed structural and morphological change in cycled samples. As shown from the SAED patterns (Figure 2a−d), after cycles, extra diffraction 976

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

Figure 2. (a−d) [100] zone SAED patterns from different samples. Red arrows in (a) highlight the C2/m featured streaks due to formation of stacking faults and local crystal rotation. Dashed red circles in (b−d) highlight extra diffraction spots due to structure transformation from C2/m to spinel. (e−h) TEM images to show particle morphology evolution from pristine sample to 45-cycles sample.

spots as highlighted by dashed red circles in each figure emerged and the intensity of the extra diffraction spots become stronger as the cycle number increased. On the other hand, the diffuse streaks (indicated by red arrows in (a)), which are due to formation of high density stacking faults and local crystal rotation in monoclinic structure with C2/m symmetry (see Supporting Information Figure S1),30,31 become weaker as the cycle number increases. The streaks on the diffraction pattern are not visible in Figure 2d, indicating little monoclinic C2/m phase left after 45 cycles. By observing from the different zone axis, we confirmed the cycling induced extra diffraction spots originating from cubic spinel-like structure (see Supporting Information Figure S2). Thus, the SAED patterns indicate the original monoclinic C2/m structure progressively transformed into a cubic spinel-like structure upon cycling. Accompanying these structural evolutions are the morphological evolutions as shown in Figure 1e−h, featuring the formation of a porous structured layer from the surface of the particle, and the thickness of this porous layer increases with increasing cycling number as further described in detail in the following section. Cycling induced structure transformation and morphological evolution were further confirmed using STEM-HAADF imaging as shown in Figure 3, illustrating the gradual morphological and structural change of the Li2MnO3 particle with the increasing number of cycles. Figure 3a,b corresponds to the pristine sample, revealing a typical monoclinic C2/m lattice structure as viewed from [100] zone. The particle surface is atomically sharp. After only one cycle, the surface of the particle shows noticeable changes as shown in Figure 3c,d, where a brighter surface layer was developed as highlighted in Figure 3d and Supporting Information Figure S3 by the red arrows. Following 10 cycles, this cycling induced surface layer becomes significantly thick and is featured by a porous structure as shown in Figure 3e,f by the clear contrast difference between inner bulk and outer surface layer at low magnification image (indicated by the dashed blue line). It is also noticed that the surface layer thickness shows dependence on the crystallographic orientations, with that on the (001) facets being slightly thinner than that on other facets (highlighted in Figure 3e). After 45 cycles, the whole particle completely lost its original morphology. As shown in Figure 3g, it seems many pits were

formed on the particle surface and well-defined particle surfaces were also damaged due to severe corrosion. The gradual propagation of the phase transformation from surface to the bulk is also confirmed by nanobeam electron diffraction on a single particle after 10 cycles. As shown in Figure 4, we scanned the particle from inner bulk to outer surface at 8 positions. The small beam size of ∼3 nm enables the collection of the local structure information with high spatial resolution. The diffraction patterns clearly show that positions 1 and 2 are pure monoclinic C2/m structure, from position 3 to position 5 it presents a mixed pattern of two phases, positions 6 and 7 are pure spinel-like structure, and the outermost position 8 shows amorphous pattern which may come from Super P carbon and/or SEI layer. Thus, it can be concluded that the C2/m→ spinel phase transformation progressively proceeds from outer surface into inner bulk upon cycling. From the nanobeam electron diffraction and previous SAED patterns (Figure 2a−d), the relative orientation between the original C2/m structure and the newly formed spinel structure is determined as [100]C2/m//[112]spinel, (001)C2/m//(111) spinel. Figure 5 shows the high resolution STEM-HAADF imaging of the 10-cycles sample from both the outer surface layer and the inner bulk. Structurally, Figure 5b,c clearly indicates the surface layer is actually not a single phase. Instead, the outermost surface layer corresponds to Mn3O4-type spinel structure (Figure 5d) and the inner surface layer is another structure, which was identified as tetragonal structure with space group as I41 (Figure 5e). Very similar surface modifications have been observed in Ni-contained LMR.32 There are two different features between the tetragonal I41 structure and the Mn3O4-spinel. First, in the I41 structure, all the Mn cations seat in octahedral sites, and there is no Mn in the tetrahedral site; as a contrast, in Mn3O4-spinel structure Mn cations seat in both octahedral and tetrahedral sites. Thus, the outmost surface Mn3O4-spinel structure is a direct spinel structure where Li cations are seated in tetrahedral sites. Second, the cation ordering is different, which results in different contrast for different Mn columns (highlighted in Figure 5d,e). The crystal structure models for tetragonal I41 structure and Mn3O4-spinel have been given in Supporting 977

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

Figure 4. Eight-position line scan of a 10-cycles sample using nanobeam electron diffraction. The crystal was tilted to the [100] zone axis.

Figure 3. STEM-HAADF images to show particle morphology and lattice structure of pristine materials (a, b), 1-cycle samples (c, d), 10cycles samples (e, f), and 45-cycles sample (g). Pristine sample shows clean and clear particle surface (b). One-cycle sample shows bright contrast on particle surface indicated by red arrows in (d). Ten-cycles sample shows clear surface layer indicated by blue dashed lines in (e) and (f). The thickness of surface layer on (001) facets is much thinner than that of other surface facets, which is highlighted by yellow markers in (e).

intensity profile scan (as shown in the bottom of Figure 5g,h). The disordered planes in a 10-cycles sample are believed to be due to the activation process. In accordance with the gradual structural change is compositional change upon cycling. Figure 6 shows EDS mapping results from a 10 cycles particle and a pristine particle. The STEM-HAADF image (Figure 6a) shows that the particle has a surface layer denoted by the dashed blue line. Figure 6b,c shows the Mn elemental and O elemental mappings, respectively. A line scan profile was extracted from the mapping results as shown in Figure 6d,e. In Figure 6d, the red arrow highlights a position showing a sharp drop of oxygen signal counts. The position is just right at the boundary between surface layer and inner bulk. To eliminate the influence of sample thickness, the O/Mn ratios were plotted in Figure 6e. It is clearly seen that the surface layer has a lower O/Mn ratio as compared with inner bulk. After subtracting the background oxygen signal (probably from the SEI layer), the averaged O/ Mn values are 2.4 ± 0.1 and 2.9 ± 0.1 for the surface layer and inner bulk, respectively. Since the standard O/Mn ratio for Li2MnO3 is 3, thus, we could estimate that 20% oxygen has been removed from the surface layer after cycling. In contrast, a pristine sample shows uniform O and Mn distributions after going through the preparation process for the electrochemical test (Figure 6f,g and Supporting Information Figure S5), which confirmed that the observed surface modification in the 10cycles samples is due to battery cycling.

Information Figure S4. Figure 5f shows another [100] zone high resolution STEM-HAADF image from a 10-cycles particle. High resolution STEM-HAADF images also show the surface layer was heavily etched, leading to the formation of high density pits and trenches. For the 10-cycles samples, the surface layer is mainly the I41 structure. For the 45-cycles samples, due to its fragile lattice structure caused by battery cycling, which can be easily damaged by electron beam at high magnification imaging, and a thick amorphous layer on particle surface, which is generally perceived to be the solid electrolyte interphase (SEI) layer, technically, it is very hard to get STEM-HAADF lattice images to identify its structure. However, according to the phase transformation sequence from surface to bulk, which is C2/m → I41 → spinel, we believe that, after prolonged cycling, the sample should correspond to a structure of LixMn3O4-spinel where x could be as low as zero. Besides the surface layer structure transformation, it has been also noted that the inner bulk lattice changes after 10 cycles. As shown in Figure 5g,h, even though both show similar C2/m monoclinic structure, the lattice ordering in a 10-cycles sample (Figure 5h) has been partially destroyed, which can be quantified by the 978

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

Figure 5. (a−c) [010] zone STEM-HAADF images to show a 10-cycles Li2MnO3 particle and its outer surface lattice structure. (d) and (e) highlight the lattice difference between Mn3O4-spinel and I41 structure. Purple and green indicate Mn and Li, respectively. (f) [100] zone STEM-HAADF image of the surface layer after 10 cycles. STEM-HAADF images to show inner bulk lattice structure of pristine sample (g) and 10-cycles sample (h). The intensity profiles along the blue lines in (g) and (h) are shown in each figure.

Mn−O phases at different Li and Mn environments, the formation energy of each phase was calculated with the Vienna ab initio simulation package. During the calculations, we considered generalized gradient approximation (GGA) with Perdew−Burke−Ernzerhof (PBE) exchange. The pseudopotentials with 2s1, 3d64s1 , and 2s22p4 valence electron configurations were used for Li, Mn, and O atoms, respectively. The formation energy per atom was estimated using Ef = (ELMO − nLiμLi − nMμM − nOμO)/n, where ELMO is the total energy of Li−Mn−O configuration; μLi, μM, and μO are chemical potentials of Li, Mn, and O referenced to metal Li, metal Mn, and the half binding energy of O2, respectively; and nLi, nM, nO, and n are the numbers of Li, Mn, O, and total atoms in the Li−Mn−O configuration, respectively, satisfying the relation of nLi + nM + nO = n. For convenience, we used Li concentration denoted by nLi/(nLi + nMn) for simulating Li and Mn environment. All calculations were carried out with spin-polarization. Electronic wave functions were expanded using a plane-wave basis set with a cutoff energy of 550 eV, and the atomic positions of the structure were relaxed until all the force components were smaller than 0.01 eV/Å. Figure 8 gives the change of formation energy with Li concentration for four different Li−Mn−O phases. The lower the formation energy, the more stable the phase is. Thus, it can be seen that with the decrease of Li concentration the Li−Mn−O phase will go through a C2/m → R3̅m → I41 → spinel phase transition. Corresponding Li concentrations for C2/m → R3̅m, R3̅m → I41, and I41 → spinel transitions are about 0.58, 0.47, and 0.30. Note that Li concentrations of phase transition here were

Besides removal of oxygen, we also confirmed the gradual depletion of Li from the very surface of the particle. As shown in Figure 7a,b, the EELS mapping of a 10-cycles particle clearly reveals that the outer surface layer has a low Li content. Such Li-depleted surface layer is consistent with the phase transformation behavior, where the Mn cations substitute the Li sites as shown in the STEM-HAADF images. A more quantitative comparison is shown in Figure 7d,e, where the EELS were collected from the positions shown in Figure 7c. From Figure 7d, it can be seen that the Li−K edge is very low for the outer surface layer, indicating the depletion of Li at the very surface. In the core loss region (Figure 6e), the Mn L2 has different intensity after normalizing the Mn L3 edge as well as the peak shift, which indicates difference of valence states of Mn cations.33,34 For the pristine sample with Mn4+ as the only valence state, the ratio of Mn L3/L2 is the lowest. For the 10cycles samples, due to the reduction of Mn4+, the ratio of Mn L3/L2 increases. This is especially true for the outer surface layer, as seen from the spectra captured at positions 1 and 2, where larger L3/L2 ratios indicate more tetravalent Mn cations being reduced in the surface layer as compared with that of the inner bulk. The valence state change is also consistent with Mn3O4-spinel where Mn3+ and Mn2+ coexist in the structure. It is known that a Mn2+ can have a high mobility, therefore favoring the C2/m to spinel transformation.35 DFT Calculations. Above experimental observation on the structural and chemical evolution of the surface layer is also consistently supported by the DFT calculation of the formation of energy of each phase. To determine relative stabilities of Li− 979

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

equally to increasing the cell’s inner resistance and results in a lower working voltage. Also, according to the Ceder et al. simulation,36 the limited channels for Li ions to diffuse through in the newly formed Mn3O4-spinel will result in a poor rate capability. Since layered LMR cathode is a composite material of Li2MnO3 and LiMO2 (M = Mn, Ni, Co et al.),11,13,30,37,38 increasing the phase homogeneity and suppressing the pure Li2MnO3 phase could improve its electrochemical performance. Comparing with the structural degradation of the Ni and/or Co contained LMR cathode in LIBs,34,39−41 the Li2MnO3 cathode shows similar structure transformation on the particle surface. However, the structural degradation of the Li2MnO3 cathode is much faster, which indicates adding Ni and Co can significantly improve LMR’s cyclic stability. Thackeray et al. proposed that it is the Ni−Mn interactions that play a key role in maintaining the average oxidation state of Mn ions above 3+ to suppress Mn migration as well as the spinel structure transformation.42 Meng et al.’s computation work shows oxygen vacancies play an important role in cation migration and structure transformation.43,44 Thus, suppressing the structure transformation can only be achieved via maintaining the high valent state (>3+) of Mn cations and low oxygen vacancy level. This has been proven in this work where we have shown Mn2+ and Mn3+ appeared and 20% oxygen has been removed in the surface phase transformation layer. We also noted the 20% oxygen removal is much higher than that of Ni contained LMR which is reported as less than 5%,39 indicating oxidation of Ni2+ to higher valent state during the charging process can also suppress oxygen removal in Ni-LMR. Another point is if many oxygen vacancies were developed during removing Li ions in the charge process, during the discharge process, due to the reinsertion of Li ions, TM cations will be reduced to a lower valent state to keep charge neutrality since oxygen can hardly be reinserted back. In addition, close packed oxygen anions are the basis of LMR crystal structure, and removing oxygen will be destructive for the crystal lattice. Thus, suppressing oxygen removal is supposed to be the most important key to improve cathode materials’ cyclic stability.

Figure 6. (a−c) EDS mapping results from a 10-cycles particle. Dashed blue line indicates the surface layer in (a). (d) A line scan signal profile extracted from the mapping results. (e) Calculated O/ Mn ratios according to the line scan results in (d). The averaged O/ Mn ratios are 2.4 and 2.9 for surface layer and inner bulk, respectively. (f) STEM-HAADF image of a pristine Li2MnO3 particle and (g) its EDS line scan results. O/Mn ratios are the same from particle surface to inner bulk.

determined at 0 K from the simulation, and thus these values of phase transition may be somewhat different to the really experimental results; for instance, we do not clearly observe the R3̅m phase in cycled samples. As shown in Figure 7b, the abrupt Li/Mn ratio change makes direct imaging of the R3̅m phase, if it exists, much more difficult. Nevertheless, the calculated results are generally consistent with the experimental observation that phase transition occurs with the change of the Li and TM environments studied in this case. Structural and Chemical Evolution Results in Electrochemical Degradation. On the basis of our above investigations, a direct correlation between structure degradation and electrochemical degradation of Li2MnO3 can be built up. For the capacity fading, in the phase transformation surface layer, at the discharge state, very little Li content is detected, which means the phase transformation surface layer is no longer a Li active layer. Li ion intercalation and deintercalation are negligible in such a structure. Therefore, after 45 cycles, the whole particle transformed into Mn3O4-spinel structure, the capacity becoming very low as shown in Figure 1a. Thus, the progressive structural transformation directly results in capacity fading by reducing the Li active volume of the particle. For voltage fading, the phase transformation layer can be contributed in two ways. First, after the original C2/m structure being transformed, the material itself will have a lower working voltage. Second, the phase transformation layer can act as a barrier layer and/or insulating layer for Li ion transportation. Thus, an increased thickness of phase transformation layer acts

■

CONCLUSION Pristine and different cycle number Li2MnO3 cathode materials have been systematically investigated using aberration corrected S/TEM. Cycling induced structure transformation was identified using electron diffraction and STEM-HAADF imaging. The structure change initiated from particle surface and propagated into inner bulk by following a certain evolution sequence during cycling: monoclinic C2/m → tetragonal I41 → cubic spinel. EDS mapping reveals that up to 20% oxygen has been removed from the surface phase transformation layer. EELS analysis indicates the surface phase transformation layer (including I41 and spinel phases) has little Li content and the tetravalent Mn cations are partially reduced to lower valence states. By combining the TEM microanalysis with electrochemical data, along with DFT calculations, we clarified the relationship between structure degradation and cell performance degradation of Li2MnO3 cathode for lithium ion batteries. In a more general term, since Li2MnO3 cathode is the parent compound for LMR cathode, this work will enhance our understanding on the degradation mechanism of LMR cathode materials during cycling. 980

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials

Figure 7. (a, b) STEM-EELS mapping of a 10-cycles sample. The mapping area is highlighted by the dashed red frame in (a). (c−e) STEM-EELS analysis of three positions (1, 2, and 3, shown in (c)) of a 10-cycles sample. The spectra are normalized using Mn−M edge and Mn−L3 edge in (d) and (e), respectively. In (d), the depressed Li−K edge in positions 1 and 2 indicates less Li content at the two positions. In (e), the depressed Mn− L2 edge was shown for the 10-cycles sample as compared with pristine sample. The lower the L2 intensity is, the more the Mn4+ cations have been reduced.

*(C.-M.W.) E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

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, Subcontract No. 6951379, under the Batteries for Advanced Transportation Technologies (BATT) Program. Part of the high resolution transmission electron microscopy study described in this paper is supported by the Laboratory Directed Research and Development Program as part of the Chemical Imaging Initiative at Pacific Northwest National Laboratory (PNNL). The work was conducted in the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by DOE’s Office of Biological and Environmental Research and located at PNNL. PNNL is operated by Battelle for the Department of Energy under Contract DE-AC05-76RLO1830.

Figure 8. Calculated lattice formation energy as a function of Li/(Li + Mn) in different structure forms (monoclinic C2/m, hexagonal R3̅m, tetragonal I41, and cubic spinel).

■

■

ASSOCIATED CONTENT

S Supporting Information *

STEM-HAADF images, SAED patterns, TEM images, EDS mapping, and crystal structural models are available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Yu, F.; Zhang, J. J.; Wang, C. Y.; Yuan, J.; Yang, Y. F.; Song, G. Z. Prog. Chem. 2010, 22, 9. (2) Li, X. L.; Meduri, P.; Chen, X. L.; Qi, W.; Engelhard, M. H.; Xu, W.; Ding, F.; Xiao, J.; Wang, W.; Wang, C. M.; Zhang, J. G.; Liu, J. J. Mater. Chem. 2012, 22, 11014. (3) Goriparti, S.; Miele, E.; De Angelis, F.; Di Fabrizio, E.; Zaccaria, R. P.; Capiglia, C. J. Power Sources 2014, 257, 421. (4) Choi, N. S.; Chen, Z.; Freunberger, S. A.; Ji, X.; Sun, Y. K.; Amine, K.; Yushin, G.; Nazar, L. F.; Cho, J.; Bruce, P. G. Angew. Chem., Int. Ed. Engl. 2012, 51, 9994.

AUTHOR INFORMATION

Corresponding Authors

*(J.-G.Z.) E-mail: [email protected]. 981

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982

Article

Chemistry of Materials (5) Tarascon, J. M.; Armand, M. Nature 2001, 414, 359. (6) Yu, H.; Zhou, H. J. Phys. Chem. Lett. 2013, 4, 1268. (7) Zheng, J.; Gu, M.; Genc, A.; Xiao, J.; Xu, P.; Chen, X.; Zhu, Z.; Zhao, W.; Pullan, L.; Wang, C.; Zhang, J. G. Nano Lett. 2014, 14, 2628. (8) Gu, M.; Genc, A.; Belharouak, I.; Wang, D. P.; Amine, K.; Thevuthasan, S.; Baer, D. R.; Zhang, J. G.; Browning, N. D.; Liu, J.; Wang, C. M. Chem. Mater. 2013, 25, 2319. (9) Johnson, C. S.; Li, N.; Lefief, C.; Vaughey, J. T.; Thackeray, M. M. Chem. Mater. 2008, 20, 6095. (10) Croy, J. R.; Kim, D.; Balasubramanian, M.; Gallagher, K.; Kang, S.-H.; Thackeray, M. M. J. Electrochem. Soc. 2012, 159, A781. (11) Bareno, J.; Balasubramanian, M.; Kang, S. H.; Wen, J. G.; Lei, C. H.; Pol, S. V.; Petrov, I.; Abraham, D. P. Chem. Mater. 2011, 23, 2039. (12) Lu, Z. H.; MacNeil, D. D.; Dahn, J. R. Electrochem. Solid-State Lett. 2001, 4, A191. (13) Yu, X.; Lyu, Y.; Gu, L.; Wu, H.; Bak, S.-M.; Zhou, Y.; Amine, K.; Ehrlich, S. N.; Li, H.; Nam, K.-W.; Yang, X.-Q. Adv. Energy Mater. 2014, 4, 1300950. (14) Zhang, X.; Luo, D.; Li, G.; Zheng, J.; Yu, C.; Guan, X.; Fu, C.; Huang, X.; Li, L. J. Mater. Chem. A 2013, 1, 9721. (15) Song, B.; Liu, H.; Liu, Z.; Xiao, P.; Lai, M. O.; Lu, L. Sci. Rep. 2013, 3, 3094. (16) Liu, W.; Fang, G.; Xia, B.; Sun, H.; Kaneko, S.; Li, D. RSC Adv. 2013, 3, 15630. (17) Yu, D. Y. W.; Yanagida, K.; Kato, Y.; Nakamura, H. J. Electrochem. Soc. 2009, 156, A417. (18) Yabuuchi, N.; Yoshii, K.; Myung, S. T.; Nakai, I.; Komaba, S. J. Am. Chem. Soc. 2011, 133, 4404. (19) Xiao, R. J.; Li, H.; Chen, L. Q. Chem. Mater. 2012, 24, 4242. (20) Bareno, J.; Lei, C. H.; Wen, J. G.; Kang, S. H.; Petrov, I.; Abraham, D. P. Adv. Mater. 2010, 22, 1122. (21) Croy, J. R.; Balasubramanian, M.; Kim, D.; Kang, S. H.; Thackeray, M. M. Chem. Mater. 2011, 23, 5415. (22) Kalyani, P.; Chitra, S.; Mohan, T.; Gopukumar, S. J. Power Sources 1999, 80, 103. (23) Thackeray, M. M. Prog. Solid State Chem. 1997, 25, 1. (24) Armstrong, A. R.; Holzapfel, M.; Novak, P.; Johnson, C. S.; Kang, S. H.; Thackeray, M. M.; Bruce, P. G. J. Am. Chem. Soc. 2006, 128, 8694. (25) Robertson, A. D.; Bruce, P. G. Chem. Mater. 2003, 15, 1984. (26) Armstrong, A. R.; Robertson, A. D.; Bruce, P. G. J. Power Sources 2005, 146, 275. (27) Reed, J.; Ceder, G.; Van der Ven, A. Electrochem. Solid-State Lett. 2001, 4, A78. (28) Rana, J.; Stan, M.; Kloepsch, R.; Li, J.; Schumacher, G.; Welter, E.; Zizak, I.; Banhart, J.; Winter, M. Adv. Energy Mater. 2014, 4, 1300998. (29) Wang, R.; He, X. Q.; He, L. H.; Wang, F. W.; Xiao, R. J.; Gu, L.; Li, H.; Chen, L. Q. Adv. Energy Mater. 2013, 3, 1358. (30) Jarvis, K. A.; Deng, Z. Q.; Allard, L. F.; Manthiram, A.; Ferreira, P. J. Chem. Mater. 2011, 23, 3614. (31) Xu, Z.; Wang, J.; Zhang, K.; Zheng, H.; Dai, Z.-X.; Gui, J.; Yang, X.-Q. ACS Appl. Mater. Interfaces 2013, 6, 1219. (32) Yan, P.; Nie, A.; Zheng, J.; Zhou, Y.; Lu, D.; Zhang, X.; Xu, R.; Belharouak, I.; Zu, X.; Xiao, J.; Amine, K.; Liu, J.; Gao, F.; ShahbazianYassar, R.; Zhang, J. G.; Wang, C. M. Nano Lett. 2015, 15, 514. (33) Wang, Z. L.; Yin, J. S.; Jiang, Y. D. Micron 2000, 31, 571. (34) Lin, F.; Markus, I. M.; Nordlund, D.; Weng, T. C.; Asta, M. D.; Xin, H. L.; Doeff, M. M. Nat. Commun. 2014, 5, 3529. (35) Kan, Y.; Hu, Y.; Lin, C. K.; Ren, Y.; Sun, Y. K.; Amine, K.; Chen, Z. Phys. Chem. Chem. Phys. 2014, 16, 20697. (36) Lee, J.; Urban, A.; Li, X.; Su, D.; Hautier, G.; Ceder, G. Science 2014, 343, 519. (37) Yu, H.; Ishikawa, R.; So, Y. G.; Shibata, N.; Kudo, T.; Zhou, H.; Ikuhara, Y. Angew. Chem., Int. Ed. Engl. 2013, 52, 5969. (38) Thackeray, M. M.; Johnson, C. S.; Vaughey, J. T.; Li, N.; Hackney, S. A. J. Mater. Chem. 2005, 15, 2257. (39) Xu, B.; Fell, C. R.; Chi, M. F.; Meng, Y. S. Energy Environ. Sci. 2011, 4, 2223.

(40) Gu, M.; Belharouak, I.; Zheng, J. M.; Wu, H. M.; Xiao, J.; Genc, A.; Amine, K.; Thevuthasan, S.; Baer, D. R.; Zhang, J. G.; Browning, N. D.; Liu, J.; Wang, C. M. ACS Nano 2013, 7, 760. (41) Bettge, M.; Li, Y.; Gallagher, K.; Zhu, Y.; Wu, Q. L.; Lu, W. Q.; Bloom, I.; Abraham, D. P. J. Electrochem. Soc. 2013, 160, A2046. (42) Kim, D.; Croy, J. R.; Thackeray, M. M. Electrochem. Commun. 2013, 36, 103. (43) Qian, D.; Xu, B.; Chi, M.; Meng, Y. S. Phys. Chem. Chem. Phys. 2014, 16, 14665. (44) Fell, C. R.; Qian, D. N.; Carroll, K. J.; Chi, M. F.; Jones, J. L.; Meng, Y. S. Chem. Mater. 2013, 25, 1621.

982

DOI: 10.1021/cm504257m Chem. Mater. 2015, 27, 975−982