Resolving the Compositional and Structural Defects of Degraded

Recently, there has been growing interest in the direct recycling of cathode ... Due to the high capacity and reduced cost, layered oxide LiNixCoyMnzO...
0 downloads 0 Views 9MB Size
Download PDF
Resolving the Compositional and Structural Defects of Degraded LiNixCoyMnzO2 Particles to Directly Regenerate High-Performance Lithium-Ion Battery Cathodes Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on June 27, 2018 at 02:51:34 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Yang Shi,† Gen Chen,‡ Fang Liu,‡ Xiujun Yue,† and Zheng Chen*,†,§ †

Department of NanoEngineering, University of California San Diego, La Jolla, California 92093, United States Department of Chemical and Biomolecular Engineering, University of California Los Angeles, Los Angeles, California 90095, United States § Program of Materials Science and Engineering, University of California San Diego, La Jolla, California 92093, United States ‡

S Supporting Information *

ABSTRACT: Layered oxide LiNixCoyMnzO2 (0 < x,y,z < 1, x + y + z = 1) or NCM is becoming the dominating cathode material in high-energy lithium-ion batteries (LIBs), which have degradation issues after cycling due to Li loss and phase changes. Directly resolving these issues to generate new cathodes cannot only reduce the high cost but also prevent environmental pollution from disposal of used LIBs. However, currently there is no effective approach to tackle this challenge. Here we demonstrate a nondestructive process to directly regenerate degraded NCM cathode particles to obtain new active particles. Using this method, nearly ideal stoichiometry, low cation mixing, and high phase purity were achieved in the regenerated NCM particles, which offer high specific capacity, good cycling stability, and high rate capability, all reaching pristine materials. Our work represents a simple yet efficient approach to directly regenerate high-performance NCM cathodes with distinct advantages over traditional hydrometallurgical methods and builds an important foundation for the sustainable manufacturing of energy materials.

L

embedded energy in the desired cathode particles is lost during such a destructive recycling process. Recently, there has been growing interest in the direct recycling of cathode materials, such as the solid-state sintering approach, in which a predetermined amount of lithium salt (e.g., Li2CO3) is sintered with the spent cathode powders to generate new particles.7,13,14 This relatively green approach simplifies the recycling process and retains the embedded energy; however, only LiCoO2 (LCO)7,13 and LiFePO414 cathode materials have been investigated. Due to the high capacity and reduced cost, layered oxide LiNixCoyMnzO2 (0 < x,y,z < 1, x + y + z = 1) or NCM is becoming the dominating cathode material in state-ofthe-art LIBs.15 So far, the recycling of NCM cathodes has been mainly based on the hydrometallurgical process, as mentioned earlier.4,16−20 Therefore, there is an urgent need to develop a

ithium-ion batteries (LIBs) have been widely used in mobile electronics, electric vehicles (EVs), and renewable grids due to their high energy density.1 Typical LIBs will reach their lifetime after a few years of service due to performance degradation. It is projected that ∼1 million tons of used LIBs will be extracted from the market by 2025.2−4 From the economic point of view, reuse of the metals from LIBs can significantly reduce their cost because more than 20% of the cost of EV batteries comes from their cathode materials.5,6 From the environmental point of view, the flammable and toxic wastes (organic solvents, heavy metals) generated from disposal of used batteries can cause severe environment pollution.7 Therefore, it becomes strongly desired to recycle, reuse, and remanufacture LIBs for sustainable energy storage.8 The state-of-the-art approach to recycle cathode materials is mainly based on a hydrometallurgical process that involves acid dissolution followed by chemical precipitation.9−12 However, large amounts of acid and base solutions are used, which generates additional waste and complicates the recycling process. More importantly, the © XXXX American Chemical Society

Received: May 21, 2018 Accepted: June 22, 2018

1683

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

Cite This: ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters

Figure 1. (a) SEM images (top: secondary particles; bottom: primary particles) and (b) size distribution of pristine NCM523 secondary particles. (c) SEM images (top: secondary particles; bottom: primary particles) and (d) size distribution of degraded NCM523 secondary particles (after 200 cycles). (e) HR-TEM image and FFT image of pristine NCM523 particles that show a pure layered phase. (f) HR-TEM image and FFT images of cycled/degraded NCM523 particles that show coexistence of layered, spinel, and rock salt phases.

from different degradation mechanisms of NCM cathode materials compared with simple LCO.22 More specifically, besides Li+ loss due to thickening of the solid −electrolyte interface (SEI),23 the change of crystal structure and microphase on the particle surface (or subsurface) is a major reason for the capacity degradation in layered oxide cathodes. For LCO, spinel phases such as Co3O4 and LiCo2O4 can form after degradation.24−26 However, in the case of NCM, the phase change is more complicated. Due to the Li+ deficiency and migration of Ni2+ between the layers, a rock salt phase (e.g., NiO) will form on the surface besides the common spinel phase.27−29 Both phases increase the charge-transfer resistance and reduce the cathode performance.22 However, reconstructing the rock salt phase into a Li+-conducting layered structure is challenging due to the thermodynamically unfavorable nature of this reaction.22

more energy-efficient, nondestructive process to directly recycle NCM cathodes. Our recent work on combining hydrothermal treatment with short thermal annealing to regenerate degraded LiCoO2 particles has demonstrated the successful reconstruction of stoichiometry composition and desired crystalline structure from severely degraded cathode materials.21 Compared with the previous approaches for cathode recycling, our new strategy represents a more effective process to regenerate faded LiCoO2 cathodes with high electrochemical performance and low energy cost. However, the cathode reactivity and stability may change dramatically with their original composition and crystal structure. The complex chemistry in the NCM cathodes can influence the change of crystal structure and local phase after cycling, which further affects the regeneration process. Accordingly, challenges may arise 1684

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters

Figure 2. (a) Illustration of the hydrothermal lithiation process in which Li+ is redosed to the Li-deficient sites to recover its desired stoichiometry. (b) Lithiation kinetics of degraded cathode particles during hydrothermal treatment. The Li+ concentration of different cathode particles changes dramatically with hydrothermal treatment temperature and time.

distribute in 3−5 μm, with primary grains 0.2−1.0 μm in diameter. The spherical morphology of the secondary particles was maintained after cycling. Because the microphase change is one of the most important reasons for cathode degradation, high-resolution transmission electron microscopic (HR-TEM) images were taken to directly observe the changes in crystal structure after cycling. More than 20 particles were examined for each sample. As expected, the pristine NCM111 (Figure S3) and NCM523 (Figure 1e) cathodes show only a layered rhombohedral structure, with the hexagonal diffraction pattern showing a (012) plane. The cycled NCM111 particles have a layered phase in the bulk region, while they show a spinel phase near the surface (Figure S4). Diffraction spots from the spinel phase were detected along with the layered phase, with the additional diffraction spots indexed as (2̅20)s.22 Similarly, diffraction spots from the spinel phase were also observed in cycled NCM523 (Figure 1f). Besides the spinel phase, the rock salt phase appears in the region of ∼3 nm from the surface of the NCM523 particles, with a reduction in the number of diffraction spots due to the high symmetry of the rock salt phase, and the diffraction spots are indexed as (2̅20)c.30 No rock salt phase was clearly detected in the cycled NCM111 cathode in our case. The phase transformation in the surface region could be due to the Li+ deficiency near the surface, which is generally observed in layered cathode materials31,32 because it is a thermodynamically favored transformation when the Li contents are reduced to half of that in the original structure.33 With the above understanding, the first step to regenerate the degraded cathode particles is to redose lithium using a hydrothermal-based solution impregnation method. Figure 2a illustrates the lithiation process of various cathode particles during the hydrothermal treatment. Interestingly, a significant difference was found for different cathode chemistries. For degraded LCO particles (Li0.8CoO2), the Li concentration can be recovered to 0.98 after being treated at 180 °C for 4 h (Figure 2b), which is not the perfect stoichiometry but commonly found for LCO even in the pristine particles.21 However, for degraded NCM111 and NCM523 particles, the Li stoichiometry cannot be fully recovered (e.g., only 0.95) after being treated at 180 °C for even 24 h. Nevertheless, once the temperature increases to 220 °C, the Li concentration can be fully recovered after treatment of only 4 h. The different lithiation kinetics of NCM and LCO may be related to the higher degree of cation mixing in NCM due to the similar sizes of Ni2+ (0.69 Å) and Li+ (0.72 Å).34 Because Ni2+ ions occupy Li+ sites, the activation energy barrier is higher for the diffusion

Here, our effort is focused on developing nondestructive approaches to directly regenerate degraded NCM cathode particles by resolving their compositional and structural defects. Specifically, we used a hydrothermal treatment combined with a short annealing step in controlled atmospheres to regenerate NCM cathode particles. As a comparison, a direct solid-state sintering approach was also examined to understand the activity of degraded NCM particles. We carefully investigated the reaction mechanism during different cathode regeneration processes. LiNi1/3Co1/3Mn1/3 O2 (NCM111) and LiNi0.5Co0.2Mn 0.3O 2 (NCM523) cathodes were selected as the model materials to study the effect of nickel content on the evolution of particle stoichiometry and microphase. With optimized conditions, both spinel and rock salt phases can be fully converted back to the layered phase using our direct regeneration approaches, as confirmed by systematic physicochemical characterizations. The lithium storage capacity and cycling stability of the degraded NCM111 and NCM523 cathode particles can be also recovered to the original levels of the pristine materials. To the best of our knowledge, this work is the first report on the direct approach to regenerate degraded NCM particles with electrochemical performance reaching brand new cathode materials. Both commercial and homemade cells were used for demonstration and performance evaluation. Cell assembly and materials harvesting for LCO and NCM111 pouch cells follow standard procedures.21 Assembly of NCM523 pouch cells is described in the experimental section (Supporting Information), with the same materials harvesting procedures as that of LCO and NCM111 pouch cells. All pouch cells were cycled in a voltage range of 3−4.5 V at 1C with capacity degradation of more than 20% (Figure S1). The degraded cathode particles were obtained and subjected to hydrothermal treatment with a short thermal annealing step (denoted as HASA) or to a direct solid-state sintering treatment by mixing with Li salt (denoted as SS). The regenerated cathode materials were made into slurries to fabricate new coin cells to evaluate their electrochemical performance for the sake of convenience and consistency (experimental procedures in the Supporting Information). Scanning electron microscopic (SEM) images and the size distribution of the pristine and NCM523 particles after 200 cycles are displayed in Figure 1a,b and c,d, respectively. The particle morphology and size distribution of the NCM111 samples were also monitored (Figure S2). All of the particles have similar morphology and sizes, in which the sizes of the secondary particles mainly 1685

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters Table 1. ICP Results of Pristine, Cycled, and Regenerated Cathode Particles sample

NCM111

NCM523

pristine cycled HT only SS-air SS-oxygen HT-SA

Li0.995Ni0.331Co0.341Mn0.329O2.012 Li0.786Ni0.328Co0.340Mn0.325O1.996 Li1.012Ni0.329Co0.341Mn0.326O2.009 Li1.016Ni0.331Co0.340Mn0.325O2.010 Li1.016Ni0.331Co0.341Mn0.326O2.012 Li1.019Ni0.330Co0.340Mn0.327O2.011

Li1.009Ni0.492Co0.209Mn0.305O2.015 Li0.788Ni0.490Co0.208Mn0.302O1.985 Li1.006Ni0.491Co0.209Mn0.304O2.012 Li1.017Ni0.490Co0.207Mn0.304O2.014 Li1.018Ni0.491Co0.208Mn0.304O2.013 Li1.021Ni0.490Co0.209Mn0.303O2.013

Figure 3. XRD patterns of pristine, cycled, and regenerated (a) NCM111 and (b) NCM523 particles by HT-SA, SS-air, and SS-oxygen approaches; enlargement of the regions in the ranges of 18.5−19.5 and 64−66° for (c) NCM111 and (d) NCM523 particles.

of Li+ because of the smaller separations between the transition metal layers.35 For the purpose of demonstration, hydrothermal treatment at 220 °C for 4 h was selected to lithiate the degraded NCM111 and NCM523 electrode particles for annealing in the next step. The compositions of different pristine, degraded, and regenerated NCM cathode materials are listed in Table 1. Compared with their pristine composition, both NCM111 and NCM523 particles had about 22% Li loss after cycling. With hydrothermal treatment, these degraded particles can be reconstituted, with the Li concentration reaching the ideal stoichiometry (∼1.0 Li). A following thermal annealing step needs to be performed to reconstruct their desired microphase and crystallinity but still maintain the Li concentration in the particles during thermal treatment. Different from the LCO cathode, in which only spinel phases (Co3O4 and LiCo2O4) are formed after cycling,24 nanoscale domains of the rock salt phase often exist in the cycled NCM cathodes besides the spinel phases.22 To convert the local rock salt MO (M = Ni, Co, Mn) domains back to layered LiMO2, the following reaction should occur36 MO + 0.5Li 2CO3 + 0.25O2 ↔ LiMO2 + 0.5CO2

This reaction indicates that oxygen partial pressure may be an important factor for the conversion process. Therefore, for comparison, the degraded NCM111 and NCM523 particles were mixed with a predetermined amount of Li+ salts to perform direct sintering to reach a target mole ratio between Li and transition metal ions (1.05:1) in both air and oxygen atmospheres. As shown in Table 1, both of the particles can reach desired overall compositions, indicating their nonsensitivity to O2 partial pressure. For the HT particles, a short annealing treatment at 850 °C for 4 h was performed in O2 to reconstruct the desired crystallinity of the material. Also, considering that possible Li loss during the annealing can lead to cation mixing,37 a small amount of excess Li was added to compensate such a Li loss. Similarly, the particles with this short annealing treatment also reached the target stoichiometry. It is more critical to investigate the evolution of the microstructure defects. The XRD patterns of pristine, degraded, and regenerated cathode particles are shown in Figure 3. A typical pattern of the α-NaFeO2 structure with R3̅m space group was observed for all samples. For both NCM111 and NCM523, the cycled cathode particles show a larger intensity ratio of I003/I104, which is consistent with those

(1) 1686

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters Table 2. Lattice Parameters of Pristine, Cycled, and Regenerated Cathode Particles sample

a /Å

c/Å

Li/Ni mixing/%

RB/%

Rwp/%

NCM111 pristine NCM111 cycled NCM111-HT-SA NCM111-SS-air NCM111-SS-oxygen NCM523 pristine NCM523 cycled NCM523-HT-SA NCM523-SS-air NCM523-SS-oxygen

2.8631(7) 2.8600(4) 2.8629(4) 2.8628(6) 2.8624(7) 2.8689(4) 2.8591(0) 2.8703(8) 2.8729(9) 2.8682(9)

14.248(3) 14.258(4) 14.246(2) 14.250(2) 14.244(7) 14.240(6) 14.319(2) 14.249(5) 14.255(9) 14.244(8)

2.42 2.81 2.07 2.45 2.43 3.39 5.10 3.74 5.53 4.28

5.92 4.9 4.22 5.61 6.14 4.41 4.92 4.99 3.93 4.96

1.47 1.52 1.43 1.58 1.51 1.65 2.53 2.17 1.71 1.75

in a previous report.22 The (003) peak shifts to lower angles, corresponding to an increase in the c lattice parameter due to electrostatic repulsion between the oxygen layers along c directions in the Li deficiency state.38 The spacing between the peaks in the (108)/(110) doublets increases after cycling, corresponding to the decrease in a lattice parameters due to the smaller effective ionic radii of Ni3+ compared to Ni2+ to compensate Li deficiency.39 After different regeneration processes, the (003) peak shifts back toward higher angles and the spacing between two doublets peaks decreases, which indicates recovery of the pristine crystal structure. Rietveld refinement was performed on all of the XRD patterns (Figures S5 and S6), and the lattice parameters are compared in Table 2. Both RB (Bragg factor) and Rwp (weighted profile R-factor) have values lower than 10%. The refinement results further confirm that the degraded particles have decreased a lattice parameters and increased c lattice parameters, clearly showing increased Li/Ni cation mixing. It is well-known that in pristine NCM cathode materials a higher I003/I104 ratio indicates less cation mixing. During cycling, the transition metals tend to migrate to the Li+ sites due to the existence of empty Li+ sites after the removal of Li+ in the charging process, which leads to accumulation of cation mixing in the cycled materials. For all of the regeneration conditions, the a and c lattice parameters change to higher and lower values, respectively. By comparing HT-SA and SS approaches, it is found that the I003/ I104 intensity ratio of the particles regenerated by the former approach is higher than that from the latter. This indicates smaller Li/Ni cation mixing of the material regenerated by the HT-SA,40 which is further confirmed by the refinement results (Table 2). It is also noted that cation mixing of the NCM111SS-air sample is similar to that of NCM111-SS-oxygen, while the cation mixing of NCM523-SS-air is larger than that of NCM523-SS-oxygen. Because a high nickel content is considered to be the key factor for the formation of the rock salt phase,41 it is speculated that the rock salt phase tends to form more easily in the cycled NCM523 cathode than that in the cycled NCM111 cathode. As the oxygen atmosphere is a critical factor that turns the rock salt phase into a layered phase, the added Li source may not effectively react with the rock salt phase when the oxygen partial pressure is low, and the migration of Ni2+ to Li+ sites continues to happen in a Li+deficient state, which leads to a higher cation mixing degree in the NCM523-SS-air. Overall, the HT-SA samples show much smaller Li/Ni mixing, suggesting its advantage of offering more homogeneous lithiation and more effective phase conversion for particle regeneration. Even though after direct regeneration no obvious changes in the morphology and particle size distribution are observed

(Figures 4a,b and S2), their microstructure needs further examination. To prove that the surface phase change can be recovered, the regenerated cathodes were carefully examined by HR-TEM (Figure 4c−e). The influence of short annealing after the hydrothermal step was investigated by comparing the images of NCM523-HT and NCM523-HT-SA samples. For NCM523-HT, there remain some amorphous domains on the surface (Figure S7), but the NCM523-HT-SA sample shows a high degree of crystallinity for each observed particle. For NCM111-HT-SA (Figure S8) and NCM523-HT-SA (Figure 4c) samples, only a layered phase in both the bulk and the surface regions was observed. For example, in the zone axis of [1̅2̅1], only a layered phase exists with the diffraction spot indexed as (012). This indicates that both the spinel and rock salt phases can be effectively converted back to the layered phase by the HT-SA approach. In addition, it is found that the change in microphase of NCM523 shows higher sensitivity to the oxygen partial pressure than NCM111 in SS regeneration. The SS in air can convert the spinel phase to the layered phase in cycled NCM111 cathode (Figure S9). However, for the NCM523 cathode, the rock salt phase can still be observed on the surface of the particles after SS in air (Figure 4d), which indicates that the oxygen partial pressure is important for such a phase conversion. By comparison, for the SS in oxygen, no spinel or rock salt phase was observed (Figure 4e), which means successful regeneration of the layered structure in the cycled NCM523 cathode. As expected, SS in oxygen can also fully recover the layered structure in the cycled NCM111 cathode as well (Figure S10). To provide further evidence of the spinel/rock salt phase in the cycled cathodes and the successful reconstruction of the layered phase after regeneration, X-ray photoelectron spectroscopy (XPS) measurement was performed on NCM111 (Figure S11) and NCM523 (Figure 4f) samples. For Ni 2p spectra, all cathodes have two dominant peaks at 854.6 eV (2p3/2) and 872.2 eV (2p1/3), which represent Ni2+, and the two less dominant shakeup peaks at 860.9 and 879.2 eV further confirm the existence of Ni2+.42 In cycled NCM111 and all NCM523 cathodes, besides the existence of Ni2+, the less prominent peaks at 857.2, 864.3, 875.5, and 882.9 eV indicate the existence of Ni3+.43 The quantitative analysis shows that the Ni2+ concentration of NCM111-cycled, NCM523-pristine, NCM523-cycled, NCM523-HT-SA, and NCM523-SS-oxygen samples are 55.14, 60.22, 72.56, 60.37, and 60.44%, respectively. The significantly higher Ni2+ concentration in the cycled NCM523 cathode is consistent with the existence of the NiO rock salt phase on the surface by TEM observation. For Mn 2p spectra, all samples have two major peaks at 642.3 1687

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters

Figure 4. (a) SEM images (top: secondary particles; bottom: primary particles) and (b) size distribution of NCM523-HT-SA particles. HRTEM and FFT images of (c) NCM523-HT-SA, (d) NCM523-SS-air, and (e) NCM523-SS-oxygen samples. Scale bars: 5 nm. (f) XPS spectra of pristine NCM523, cycled NCM523, NCM523-HT-SA, and NCM523-SS-oxygen samples.

eV (2p3/2) and 653.8 eV (2p1/3), which represent Mn4+.42 The peaks at 640.9 and 652.4 eV found in cycled NCM111 and NCM523 samples indicate the existence of Mn3+,44 while such peaks were not observed in pristine and regenerated samples. Quantitative analysis of the Mn 2p spectrum shows that the

Mn3+ concentration is 34.58 and 36.54% in cycled NCM111 and NCM523 cathodes, respectively. This is expected because the layered spinel/rock salt transformation originates from oxygen loss, which results in the formation of Mn3+ for charge compensation.45 Therefore, the existence of Mn3+ in cycled 1688

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters

Figure 5. (a) Cycling performance of pristine, nontreated, and regenerated NCM111 samples at 1C; HT-SA: hydrothermal treatment at 220 °C for 4 h, followed by annealing at 850 °C for 4 h; SS-air: sintering at 850 °C for 12 h in air; SS-oxygen: sintering at 850 °C for 12 h in oxygen. (b) Cycling performance of pristine, nontreated, and regenerated NCM523 samples at 1C. (c) Rate performance of NCM111 samples. (d) Rate performance of NCM523 samples. (e) Voltage profiles of NCM111 samples at 5C. (f) Voltage profiles of NCM523 samples at 5C. (g) Illustration of the crystal structure change of NCM523 after cycling and regeneration. The right scheme shows the atomic arrangement of layered, spinel, and rock salt phases along the [1̅2̅1] zone axis (same as TEM images).

the specific capacity and capacity retention of the regenerated cathode, it cannot fully recover the cycling performance. This is also consistent with the TEM observation (Figure 4d), which shows the rock salt phase remaining on the surface. The rate capability and voltage profiles of the pristine and regenerated cathodes at 5C are compared in Figure 5c−f. For both NCM111 and NCM523 samples, the HT-SA approach delivers better rate capability (Figure 5c,d), as well as smaller voltage drops and polarization (Figure 5e,f) than the SS approach. The rate capability of the NCM523-SS-oxygen sample is better than that of the NCM523-SS-air sample (Figure 5d), and the voltage drops and polarization of the former are smaller than those of the latter. The reason for the better rate capability of HT-SA samples is believed to relate to their lower Li/Ni mixing (Table 2). It has also been reported that hydrothermal treatment is effective in suppressing Li/Ni mixing.46 Cation mixing blocks the Li+ transportation channel and therefore decreases the rate capability.47 The different rate performance of NCM111-SS-air and NCM523-SS-air samples matches well with their microphase structures (Figures S9 and 4d). The layered structure of NCM111-SS-air is recovered, while in NCM523-SS-air, the rock salt phase still exists, which results in the inhibition of Li+ motion. To further understand the rate performance, EIS measurement was performed on pristine and regenerated NCM523 cathodes after 100 cycles at 1C (Figure S12). As shown in Table S1, the NCM523-SS-air sample has much larger charge-

cathodes and its disappearance in regenerated cathodes further support that the spinel and rock salt phases formed after cycling were recovered in a layered structure after regeneration. The electrochemical performance of the NCM111 and NCM523 cathode particles in different conditions was evaluated in a voltage range of 3−4.3 V at 1C (C = 150 mA g−1) after one activation cycle at C/10 (Figure 5). For NCM111, the pristine cathode shows a capacity of 145.1 mAh g−1 in the first cycle at 1C and 123.8 mAh g−1 after 100 cycles (Figure 5a). The nontreated, cycled cathode shows a capacity of 98.4 mAh g−1 after 100 cycles, which is due to the existence of the spinel phase on the surface (Figure S4). The electrochemical activity of the regenerated cathodes from HT-SA, SS in air, and SS in oxygen was fully recovered, with a capacity of 158.4, 153.3, and 157.4 mAh g−1 in the first cycle at 1C and 122.6, 125.4, and 123.8 mAh g−1 after 100 cycles, respectively. For NCM523, the pristine cathode shows a capacity of 146.6 mAh g−1 in the first cycle at 1C and 130.4 mAh g−1 after 100 cycles (Figure 5b). The nontreated, cycled cathode shows poor cycling performance with only 88.6 mAh g−1 capacity retention after 100 cycles. The fast capacity decay is consistent with the observed spinel and rock salt phases at the surface (Figure 1f). After regeneration by HT-SA and SS in oxygen, the cycling stability of both cathodes was fully recovered and the regenerated cathodes could maintain a capacity of 128.3 mAh g−1 (HT-SA) and 127.4 mAh g−1 (SS) after 100 cycles. However, even though the SS in air increases 1689

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

ACS Energy Letters

■

transfer resistance (Rct) (367.4 Ω) than the NCM523-SSoxygen sample (198.7 Ω); the NCM523-HT-SA sample has the smallest Rct (142.8 Ω). The superior rate performance of the HT-SA sample is attributed to its lowest Rct which favors the charge-transfer reaction for Li+ intercalation.48 The linear part of the Nyquist plot in the low-frequency range is directly related to Li+ diffusion in the electrode, and the diffusion coefficient (DLi+) could be calculated by the EIS method using Warburg impedance (Table S1).49,50 Three cells were measured for each type of NMC particles to reduce the variation of DLi+ values. The NCM523-HT-SA samples have the largest DLi+, with an average value of 4.70 × 10−12 cm2 s−1 and a standard deviation (SD) of 0.350 × 10−12 cm2 s−1. The NCM523-SS-oxygen samples have larger DLi+ values (average 1.24 × 10−12 cm2 s−1, SD 0.119 × 10−12 cm2 s−1) than the NCM523-SS-air samples (average 5.60 × 10−13 cm2 s−1, SD 0.617 × 10−13 cm2 s−1). The lower Rct and higher DLi+ of the regenerated cathode by the HT-SA approach explain its better rate capability than the cathode regenerated by the SS approach. The remaining rock salt phase and cation mixing on the surface of the regenerated NCM523 in air (Figure 4d) leads to large Rct and low DLi+. It is noted that DLi+ values of the HT-SA samples are slightly higher than those of the pristine samples (average 3.70 × 10−12 cm2 s−1, SD 0.422 × 10−12 cm2 s−1), which is probably due to (1) the cell variations and/or (2) the elimination of defects in pristine materials that possibly formed during the large-scale manufacturing process. The exact reasons will be investigated in more details for our future studies by comparing different commercial cells. Overall, degraded NCM particles with different compositional deficiencies and microphase impurities can be effectively regenerated using our direct methods that combine hydrothermal lithiation and short annealing, which leads to ideal stoichiometry, low cation mixing, and high phase purity. Figure 5g simply illustrates such a reversible process for particles after cycling and regeneration, using the NCM523 cathode as the example. After extensive cycling, a scattered rock salt phase forms at the surface of the NCM523 cathode together with the spinel phase accompanying Li loss. With direct regeneration, the undesired spinel and rock phases are converted back to the layered phase with lithiation and thermal annealing. In conclusion, we demonstrated the successful regeneration of the chemical composition and microstructure of degraded NCM cathodes using a novel particle-to-particle approach. Particularly, the perfect reconstruction of stoichiometry and microphase purity enabled by hydrothermal treatment with short annealing provided the regenerated NCM cathode particles with high capacity, good cycling stability, and high rate performance, all reaching their pristine materials, even with high Ni content. The direct solid-state sintering in air can restore the cycling stability of NCM111 but not that of NCM523 due to the higher nickel content in the latter type of cathode. The oxygen partial pressure needs to be kept high to effectively convert the rock salt phase impurities to the layered phase in NCM cathodes with high Ni content. Our work represents a simple yet efficient approach to directly regenerate high-performance NCM cathodes, with distinct advantages over traditional hydrometallurgical methods. This study also builds an important foundation for sustainable manufacturing of energy materials.

Letter

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsenergylett.8b00833. Experimental section and supplementary figures, including SEM images, TEM images, Rietveld refinement of the XRD patterns, XPS patterns, and EIS measurement (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Yang Shi: 0000-0003-2002-0343 Gen Chen: 0000-0003-3504-3572 Zheng Chen: 0000-0002-9186-4298 Notes

The authors declare the following competing financial interest(s): Z.C. and Y.S. report a patent application published on Jan 5, 2018, Systems and Methods for Regeneration of Lithium Cathode Materials, U.S. Provisional Application Serial No. 62/614300.

■ ■

ACKNOWLEDGMENTS Z.C. acknowledges start-up fund support from the Jacob School of Engineering at UC San Diego. REFERENCES

(1) Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D. Challenges in the Development of Advanced Li-Ion Batteries: A Review. Energy Environ. Sci. 2011, 4, 3243−3262. (2) Zheng, X.; Gao, W.; Zhang, X.; He, M.; Lin, X.; Cao, H.; Zhang, Y.; Sun, Z. Spent Lithium-Ion Battery Recycling-Reductive Ammonia Leaching of Metals from Cathode Scrap by Sodium Sulphite. Waste Manage. 2017, 60, 680−688. (3) Patel, P.; Ellis, T.; Howes, J. How Green is Your Electric Vehicle? MRS Bull. 2017, 42, 416−417. (4) Heelan, J.; Gratz, E.; Zheng, Z.; Wang, Q.; Chen, M.; Apelian, D.; Wang, Y. Current and Prospective Li-Ion Battery Recycling and Recovery Processes. JOM 2016, 68, 2632−2638. (5) Chen, X.; Luo, C.; Zhang, J.; Kong, J.; Zhou, T. Sustainable Recovery of Metals from Spent Lithium-Ion Batteries: A Green Process. ACS Sustainable Chem. Eng. 2015, 3, 3104−3113. (6) Ahmed, S.; Nelson, P. A.; Gallagher, K. G.; Susarla, N.; Dees, D. W. Cost and Energy Demand of Producing Nickel Manganese Cobalt Cathode Material for Lithium Ion Batteries. J. Power Sources 2017, 342, 733−740. (7) Chen, S.; He, T.; Lu, Y.; Su, Y.; Tian, J.; Li, N.; Chen, G.; Bao, L.; Wu, F. Renovation of LiCoO2 with Outstanding Cycling Stability by Thermal Treatment with Li2CO3 from Spent Li-Ion Batteries. J. Energy Storage 2016, 8, 262−273. (8) Mantuano, D. P.; Dorella, G.; Elias, R. C. A.; Mansur, M. B. Analysis of a Hydrometallurgical Route to Recover Base Metals from Spent Rechargeable Batteries by Liquid-Liquid Extraction with Cyanex 272. J. Power Sources 2006, 159, 1510−1518. (9) Li, L.; Dunn, J. B.; Zhang, X. X.; Gaines, L.; Chen, R. J.; Wu, F.; Amine, K. Recovery of Metals from Spent Lithium-Ion Batteries with Organic Acids as Leaching Reagents and Environmental Assessment. J. Power Sources 2013, 233, 180−189. (10) Iizuka, A.; Yamashita, Y.; Nagasawa, H.; Yamasaki, A.; Yanagisawa, Y. Separation of Lithium and Cobalt from Waste Lithium-Ion Batteries via Bipolar Membrane Electrodialysis Coupled with Chelation. Sep. Purif. Technol. 2013, 113, 33−41.

1690

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters (11) Zou, H.; Gratz, E.; Apelian, D.; Wang, Y. A Novel Method to Recycle Mixed Cathode Materials for Lithium Ion Batteries. Green Chem. 2013, 15, 1183−1191. (12) Jha, M. K.; Kumari, A.; Jha, A. K.; Kumar, V.; Hait, J.; Pandey, B. D. Recovery of Lithium and Cobalt from Waste Lithium Ion Batteries of Mobile Phone. Waste Manage. 2013, 33, 1890−1897. (13) Nie, H.; Xu, L.; Song, D.; Song, J.; Shi, X.; Wang, X.; Zhang, L.; Yuan, Z. LiCoO2: Recycling from Spent Batteries and Regeneration with Solid State Synthesis. Green Chem. 2015, 17, 1276−1280. (14) Li, X.; Zhang, J.; Song, D.; Song, J.; Zhang, L. Direct Regeneration of Recycled Cathode Material Mixture from Scrapped LiFePO4 Batteries. J. Power Sources 2017, 345, 78−84. (15) Shi, Y.; Zhang, M.; Qian, D.; Meng, Y. S. Ultrathin Al2O3 Coatings for Improved Cycling Performance and Thermal Stability of LiNi0.5Co0.2Mn0.3O2 Cathode Material. Electrochim. Acta 2016, 203, 154−161. (16) Sa, Q.; Gratz, E.; He, M.; Lu, W.; Apelian, D.; Wang, Y. Synthesis of High Performance LiNi1/3Mn1/3Co1/3O2 from Lithium Ion Battery Recovery Stream. J. Power Sources 2015, 282, 140−145. (17) Sa, Q.; Gratz, E.; Heelan, J. A.; Ma, S.; Apelian, D.; Wang, Y. Synthesis of Diverse LiNixMnyCozO2 Cathode Materials from Lithium Ion Battery Recovery Stream. J. Sustainable Metall. 2016, 2, 248−256. (18) Li, L.; Fan, E.; Guan, Y.; Zhang, X.; Xue, Q.; Wei, L.; Wu, F.; Chen, R. Sustainable Recovery of Cathode Materials from Spent Lithium-Ion Batteries Using Lactic Acid Leaching System. ACS Sustainable Chem. Eng. 2017, 5, 5224−5233. (19) Li, L.; Bian, Y.; Zhang, X.; Guan, Y.; Fan, E.; Wu, F.; Chen, R. Process for Recycling Mixed-Cathode Materials from Spent LithiumIon Batteries and Kinetics of Leaching. Waste Manage. 2018, 71, 362− 371. (20) Li, L.; Bian, Y.; Zhang, X.; Xue, Q.; Fan, E.; Wu, F.; Chen, R. Economical Recycling Process for Spent Lithium-Ion Batteries and Macro- and Micro-scale mechanistic study. J. Power Sources 2018, 377, 70−79. (21) Shi, Y.; Chen, G.; Chen, Z. Effective Regeneration of LiCoO2 from Spent Lithium-Ion Batteries: A Direct Approach towards HighPerformance Active Particles. Green Chem. 2018, 20, 851−862. (22) Jung, S. K.; Gwon, H.; Hong, J.; Park, K. Y.; Seo, D. H.; Kim, H.; Hyun, J.; Yang, W.; Kang, K. Understanding the Degradation Mechanisms of LiNi0.5Co0.2Mn0.3O2 Cathode Material in Lithium Ion Batteries. Adv. Energy Mater. 2014, 4, 1300787. (23) Verma, P.; Maire, P.; Novák, P. A Review of the Features and Analyses of the Solid Electrolyte Interphase in Li-Ion Batteries. Electrochim. Acta 2010, 55, 6332−6341. (24) Hausbrand, R.; Cherkashinin, G.; Ehrenberg, H.; Gröting, M.; Albe, K.; Hess, C.; Jaegermann, W. Fundamental Degradation Mechanisms of Layered Oxide Li-Ion Battery Cathode Materials: Methodology, Insights and Novel Approaches. Mater. Sci. Eng., B 2015, 192, 3−25. (25) Gabrisch, H.; Yazami, R.; Fultz, B. Hexagonal to Cubic Spinel Transformation in Lithiated Cobalt Oxide: TEM Investigation. J. Electrochem. Soc. 2004, 151, A891−A897. (26) Yano, A.; Shikano, M.; Ueda, A.; Sakaebe, H.; Ogumi, Z. LiCoO2 Degradation Behavior in the High-Voltage Phase Transition Region and Improved Reversibility with Surface Coating. J. Electrochem. Soc. 2017, 164, A6116−A6122. (27) 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. 2015, 4, 5694. (28) Myung, S. T.; Maglia, F.; Park, K. J.; Yoon, C. S.; Lamp, P.; Kim, S. J.; Sun, Y. K. Nickel-Rich Layered Cathode Materials for Automotive Lithium-Ion Batteries: Achievements and Perspectives. ACS Energy Lett. 2017, 2, 196−223. (29) Zheng, J.; Yan, P.; Zhang, J.; Engelhard, M. H.; Zhu, Z.; Polzin, B. J.; Trask, S.; Xiao, J.; Wang, C.; Zhang, J. Suppressed Oxygen Extraction and Degradation of LiNixMnyCozO2 Cathodes at High Charge Cut-off Voltages. Nano Res. 2017, 10, 4221−4231. (30) Wu, L.; Nam, K. W.; Wang, X.; Zhou, Y.; Zheng, J. C.; Yang, X. Q.; Zhu, Y. Structural Origin of Overcharge-Induced Thermal

Instability of Ni-Containing Layered-Cathodes for High-EnergyDensity Lithium Batteries. Chem. Mater. 2011, 23, 3953−3960. (31) Kobayashi, H.; Shikano, M.; Koike, S.; Sakaebe, H.; Tatsumi, K. Structural Origin of Overcharge-Induced Thermal Instability of NiContaining Layered-Cathodes for High-Energy-Density Lithium Batteries. J. Power Sources 2007, 174, 380−386. (32) Nanda, J.; Remillard, J.; O’Neill, A.; Bernardi, D.; Ro, T.; Nietering, K. E.; Go, J. Y.; Miller, T. J. Local State-of-Charge Mapping of Lithium-Ion Battery Electrodes. Adv. Funct. Mater. 2011, 21, 3282− 3290. (33) Ceder, G.; Van der Ven, A. Phase Diagrams of Lithium Transition Metal Oxides: Investigations from First Principles. Electrochim. Electrochim. Acta 1999, 45, 131−150. (34) Shannon, R. Revised Effective Ionic Radii and Systematic Studies of Interatomic Distances in Halides and Chalcogenides. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1976, 32, 751−767. (35) Zou, Y.; Yang, X.; Lv, C.; Liu, T.; Xia, Y.; Shang, L.; Waterhouse, G. I. N.; Yang, D.; Zhang, T. Multishelled Ni-Rich Li(NixCoyMnz)O2 Hollow Fibers with Low Cation Mixing as HighPerformance Cathode Materials for Li-Ion Batteries. Adv. Sci. 2017, 4, 1600262. (36) Liu, H.; Yang, Y.; Zhang, J. Reaction Mechanism and Kinetics of Lithium Ion Battery Cathode Material LiNiO2 with CO2. J. Power Sources 2007, 173, 556−561. (37) Yang, K.; Fan, L. Z.; Guo, J.; Qu, X. Significant Improvement of Electrochemical Properties of AlF3-Coated LiNi0.5Co0.2Mn0.3O2 Cathode Materials. Electrochim. Acta 2012, 63, 363−368. (38) Mohanty, D.; Kalnaus, S.; Meisner, R. A.; Rhodes, K. J.; Li, J.; Payzant, E. A.; Wood, D. L.; Daniel, C. Structural Transformation of a Lithium-Rich Li1.2Co0.1Mn0.55Ni0.15O2 Cathode during High Voltage Cycling Resolved by in situ X-ray Diffraction. J. Power Sources 2013, 229, 239−248. (39) Mohanty, D.; Gabrisch, H. Microstructural Investigation of LixNi1/3Mn1/3Co1/3O2 (x ≤ 1) and Its Aged Products via Magnetic and Diffraction Study. J. Power Sources 2012, 220, 405−412. (40) Zhang, X.; Mauger, A.; Lu, Q.; Groult, H.; Perrigaud, L.; Gendron, F.; Julien, C. M. Synthesis and Characterization of LiNi1/3Mn1/3Co1/3O2 by Wet-Chemical Method. Electrochim. Acta 2010, 55, 6440−6449. (41) Zhang, H.; Omenya, F.; Yan, P.; Luo, L.; Whittingham, M. S.; Wang, C.; Zhou, G. Rock-Salt Growth-Induced (003) Cracking in a Layered Positive Electrode for Li-Ion Batteries. ACS Energy Lett. 2017, 2, 2607−2615. (42) Li, J.; Xiong, S.; Liu, Y.; Ju, Z.; Qian, Y. Uniform LiNi1/3Co1/3Mn1/3O2 Hollow Microspheres: Designed Synthesis, Topotactical Structural Transformation and Their Enhanced Electrochemical Performance. Nano Energy 2013, 2, 1249−1260. (43) Chen, Z.; Wang, J.; Chao, D.; Baikie, T.; Bai, L.; Chen, S.; Zhao, Y.; Sum, T. C.; Lin, J.; Shen, Z. Hierarchical Porous LiNi1/3Co1/3Mn1/3O2 Nano-/Micro Spherical Cathode Material: Minimized Cation Mixing and Improved Li+ Mobility for Enhanced Electrochemical Performance. Sci. Rep. 2016, 6, 25771. (44) Tan, B. J.; Klabunde, K. J.; Sherwood, P. M. A. XPS Studies of Solvated Metal Atom Dispersed (SMAD) Catalysts. Evidence for Layered Cobalt-Manganese Particles on Alumina and Silica. J. Am. Chem. Soc. 1991, 113, 855−861. (45) Ben, L.; Yu, H.; Chen, B.; Chen, Y.; Gong, Y.; Yang, X.; Gu, L.; Huang, X. Unusual Spinel-to-Layered Transformation in LiMn2O4 Cathode Explained by Electrochemical and Thermal Stability Investigation. ACS Appl. Mater. Interfaces 2017, 9, 35463−35475. (46) Chen, Y.; Li, P.; Li, Y.; Su, Q.; Xue, L.; Han, Q.; Cao, G.; Li, J. Enhancing the High-Voltage Electrochemical Performance of the LiNi0.5Co0.2Mn0.3O2 Cathode Materials via Hydrothermal Lithiation. J. Mater. Sci. 2018, 53, 2115−2126. (47) Liu, B. S.; Wang, Z. B.; Yu, F. D.; Xue, Y.; Wang, G. J.; Zhang, Y.; Zhou, Y. X. Facile Strategy of NCA Cation Mixing Regulation and Its Effect on Electrochemical Performance. RSC Adv. 2016, 6, 108558−108565. 1691

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692

Letter

ACS Energy Letters (48) Ogihara, N.; Itou, Y.; Sasaki, T.; Takeuchi, Y. Impedance Spectroscopy Characterization of Porous Electrodes under Different Electrode Thickness Using a Symmetric Cell for High-Performance Lithium-Ion Batteries. J. Phys. Chem. C 2015, 119, 4612−4619. (49) Xia, L.; Qiu, K. H.; Gao, Y. Y.; He, X.; Zhou, F. D. High Potential Performance of Cerium-Doped LiNi0.5Co0.2Mn0.3O2 Cathode Material for Li-Ion Battery. J. Mater. Sci. 2015, 50, 2914−2920. (50) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, 2nd ed.; Wiley: New York, 2001.

1692

DOI: 10.1021/acsenergylett.8b00833 ACS Energy Lett. 2018, 3, 1683−1692