Effect of Secondary Heat Treatment after a Washing on the Electrochemical Performance of Co-Free LiNi0.975Al0.025O2 Cathodes for Li-Ion Batteries

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Effect of Secondary Heat Treatment after a Washing on the Electrochemical Performance of Co-Free LiNi0.975Al0.025O2 Cathodes for Li-Ion Batteries © 2023 The Authors. Small published by Wiley-VCH GmbH Published version Välikangas, Juho; Laine, Petteri; Hu, Tao; Tynjälä, Pekka; Selent, Marcin; Molaiyan, Palanivel; Jürgen, Kahr; Lassi, Ulla Välikangas, J., Laine, P., Hu, T., Tynjälä, P., Selent, M., Molaiyan, P., Jürgen, K., & Lassi, U. (2024). Effect of Secondary Heat Treatment after a Washing on the Electrochemical Performance of CoFree LiNi0.975Al0.025O2 Cathodes for Li-Ion Batteries. Small, 20(4), Article 2305349. https://doi.org/10.1002/smll.202305349 2024 RESEARCH ARTICLE www.small-journal.com Effect of Secondary Heat Treatment after a Washing on the Electrochemical Performance of Co-Free LiNi0.975Al0.025O2 Cathodes for Li-Ion Batteries Juho Välikangas,* Petteri Laine, Tao Hu, Pekka Tynjälä, Marcin Selent, Palanivel Molaiyan,* Kahr Jürgen, and Ulla Lassi* The steadily growing electric vehicle market is a driving force in low-cost, high-energy-density lithium-ion battery development. To meet this demand, LiNi0.975Al0.025O2(LNA), a high-energy-density and cobalt-free cathode material, has been developed using a low-cost and efficient co-precipitation and lithiation process. This article explores how further processing (i.e., washing residual lithium from the secondary particle surface and applying a secondary heat treatment at 650 °C) changes the chemical environment of the surface and the electrochemical performance of the LNA cathode material. After washing, a nonconductive nickel oxide (NiO) phase is formed on the surface, decreasing the initial capacity in electrochemical tests, and suppressing high-voltage (H2) to (H3) phase transition results in enhanced cycle properties. Furthermore, the secondary heat treatment re-lithiates surface NiO back to LNAand increases the initial capacity with enhanced cycle properties. Electrochemical tests are performed with the cells without tap charge to suppress the H2 to H3 phase transition. Results reveal that avoiding charging cells at a high voltage for a long time dramatically improves LNA’s cycle life. In addition, the gas analysis tests performed during charge and discharge to reveal how the amount of residual lithium compounds on the surface affects gas formation are studied. 1. Introduction Lithium cobalt oxide (LiCoO2)[1] has long been the most used cathode material in lithium-ion batteries (LIBs). This material J. Välikangas, P. Laine, T. Hu, P. Tynjälä, P. Molaiyan, U. Lassi Research Unit of Sustainable Chemistry University of Oulu P.O. Box 4000, Oulu FI-90014, Finland E-mail: juho.valikangas@oulu.fi;[email protected]; ulla.lassi@oulu.fi The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/smll.202305349 © 2023 The Authors. Small published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. DOI: 10.1002/smll.202305349 is especially popular for portable electronics because of its high energy density. Increased demand for LIBs, especially for use in electric vehicles, has been the driving force in efforts to replace rare and expensive cobalt in cathodes.[2] Cobalt is most replaced with nickel and manganese in LiNix−y−zCoyMnxO2 (NCM) or with nickel and aluminum in LiNix−y−zCoyAlxO2(NCA); however, a wide range of metal ratios are produced.[3] High-nickel cathodes are attractive because the increased amount of nickel in the structure increases the specific capacity of the cathode material.[4,5] However, there are several problems with this ratio, including Li/Ni mixing during synthesis, poor electrochemical cycling stability, and thermal instability.[6–10] During the high-temperature lithiation process, excess lithium is used to suppress Li/Ni mixing and compensate for lithium evaporation during the synthesis. Lithium excess led to high amount of residual lithium compounds, such as lithium carbonate (Li2CO3) and lithium hydroxide (LiOH), on the secondary particle surface.[11] Moreover, the amount of Li2CO3 on the high-nickel cathode surface could increase after synthesis if the surface is exposed to air. Since Ni3+has lower chemical stability than Co3+, the high-nickel cathode surface is more sensitive J. Välikangas, P. Laine, P. Tynjälä, U. Lassi Applied Chemistry University of Jyvaskyla, Kokkola University Consortium Chydenius Talonpojankatu 2B, Kokkola FI-67100, Finland M. Selent Centre for Material Analysis University of Oulu P.O. Box 4000, Oulu FI-90014, Finland P. Molaiyan, K. Jürgen AIT Austrian Institute of Technology GmbH Center for Low-Emission Transport Battery Technologies Giefinggasse 2, Vienna 1210, Austria Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (1 of 11) www.advancedsciencenews.com www.small-journal.com Figure 1. a) A schematic of cobalt-free LNA cathode synthesis via Co-precipitation and solid-state method, b) XRD patterns of the pristine (LNA), washed (LNA-w), and secondary heat-treated (LNA-w-650) samples. to moisture and carbon dioxide. As such, lithium (Li) and oxygen (O2) from the structure react with carbon dioxide (CO2), forming thin layers of NiO and Li2CO3on the surface. The amount of residual lithium compounds on the surface is critical for electrochemical performance. The compounds can react with the electrolytes during the electrochemical reaction and accelerate the formation of hydrofluoric acid (HF), which, in turn, can form insulating layers and decrease electrochemical performance. Furthermore, a high amount of electrolyte decomposing gases can lead to cell swelling, bad electrochemical performance, and cell failure. In addition, moisture-sensitive residual lithium compounds can lead to cathode slurry gelation and difficulties in the electrode fabrication process.[12–18] In a previous paper, we showed that, compared to pure lithium nickel oxide (LiNiO2), aluminum-modified LNA is an attractive candidate for next-generation cathode material in LIBs because it has a high capacity (215 mAh g−1) at a reasonable voltage range (4.2–2.6 V) and moderate cycling stability after washing process.[19] During the washing, most residual Li2CO3and LiOH compounds that originated from the lithiation process are washed from the surface, and a thin layer of delithiated phase NiO forms on the surface of LiNiO2. The washing process is more challenging for the pure LiNiO2than those containing cobalt because Ni3+has lower chemical stability than Co3+.[12] Washing and secondary heat treatment increase the manufacturing costs of the cathode material. However according to Ahmed et al., for NCM cathode material annual costs of purchased capital equipment are 1.4% and utilities (Electricity, Natural gas, and water) are only 1%, however same time raw material costs are 54%. Price of the cobalt is much higher than nickel so it’s obvious that minimizing the amount of cobalt even with new process steps is economically beneficial.[20] In this work, the effects of washing on the chemical structure on the material surface determine how that will affect electrochemical performance and gas formation during the cycling tests. After washing, a secondary heat treatment is performed with the aim to restore the surface structure and the electrochemical performance. Yoon et al. show that H2 to H3 phase transition could be avoided by lowering the charge cutoff voltage.[21] Recently Guo et al. show that electrolytes with low surface reactivity with LNO at a high state of charge can increase the cycle life of LNO/Li cells.[3] Furthermore, our studies show that modifying the charging protocol was identified as an effective way to suppress the H2 to H3 phase transition and dramatically improve capacity retention in full cell tests. 2. Results and Discussion 2.1. Effect of Washing and Secondary Heat Treatment of LNA Cathode Material 2.1.1. Structural and Morphology Investigations of LNA XRD patterns for LNA cathodes are shown in Figure 1b. They indicated that the pristine (LNA), washed (LNA-w), and secondary heat-treated (LNA-w-650) samples had an 𝛼-NaFeO2-type structure that was similar to that of the R3m space group (ICDD: 04023-9746). Small peaks indexed for Li2CO3were detected only for unwashed LNA shown in Figure 1b and Figure S1a (Supporting Information). The lattice parameters and Ni occupations on the Li site obtained from the XRD data are shown in Table 1 and Table S1 (Supporting Information). The lattice parameters for LNA-w increased slightly; however, these changes might be due to the removal of residual lithium and impurities from the surfaces as a Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (2 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Figure 2. FESEM images of a,b) pristine LNA, c,d) washed LNA-w, and e,f) wash and heat-treated LNA-w-650. The FESEM morphology at 10 um and 200 nm. result of washing. In other words, washing may not have changed the unit size of the material. XPS analyses revealed an increased amount of NiO phase on the LNA-w surface, which could have slightly affected the lattice parameters. For LNA-w-650, lattice parameters slightly increase, and the amount of nickel increase in the Li site, indicating that a lower amount of lithium on the second heat-treatment increase the disorder in the layered structure. Table 1. Cell parameters of the pristine (LNA), washed (LNA-w), and secondary heat-treated (LNA-w-650) samples. The error in lattice parameter 2.87536(2) means ±0.00002. Parametersa) Unit LNA LNA-w LNA-w-650 c-axis Å 14.1983(2) 14.2011(2) 14.2059(2) a-axis Å 2.87536(2) 2.87552(2) 2.87621(2) c/3a1.6460 1.6462 1.6464 Crystallite size a Å 1802.57 1935.15 2201.84 Crystallite size c Å 2516.04 2681.92 2902.31 Ni on Li site 0.0220 0.0250 0.032 As expected, the crystal size also increased during secondary heat treatment. As shown in Figure 2, all prepared materials had a roundshaped secondary particle morphology that was about 8 μm diameter. Moreover, 50 000 magnifications revealed that the secondary particles consisted of primary particles of about 200 nm. Washing and secondary heat treatment did not affect particle morphology, indicating that impurities and different amounts of residual lithium compounds formed a few nanometer-thin layers on the particle surface that could not be detected in FESEM images. Kim et al. and Zhuang et al. reported similar observations for the residual lithium compounds on the NCA type cathode material.[13,14] As shown in Table 2, the ICP-OES results revealed that sulfur, sodium, and other impurities were present on the unwashed material surface and effectively removed by the washing process. Titration analysis (Li wt%) and the Li/Me ratio measured using ICP-OES revealed that washing removed most residual Li2CO3and LiOH compounds from the secondary particle surface. There was a low level of residual lithium compounds in LNA-w. LNA-w-650, which had a secondary heat treatment at 650 °CinanO 2atmosphere, had a higher concentration Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (3 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Table 2. Li2CO3, LiOH, and Li concentrations on particle surfaces (detected by titration analysis) and Li, Ni, Co, Al, Mn, S, and Na concentrations and the Li/Me ratio (detected by ICP-OES). Sample Li2CO3LiOH Li Ni Co Al S Na Li/Me [wt%] [wt%] [wt%] [mol%] [mol%] [mol%] [mg g−1][mgg −1] Ratio LNA 1.65 0.85 0.56 97.5 0.016 2.5 1.87 0.14 1.073 LNA-w 0.46 0.01 0.09 97.4 0.016 2.6 0.14 0.05 1.020 LNA-w-650 0.44 0.35 0.18 97.4 0.015 2.5 0.14 0.05 1.034 of residual lithium compounds, indicating that heat treatment accelerated the formation of Li2CO3and LiOH on the surface. Xiong et al. obtained similar post-washing results when studying secondary heat-treated LiNi0.8Co0.1Mn0.1O2(NMC811) cathode material.[22] Surface compounds and oxidation states were further analyzed using XPS. Figure 3a–c shows the O1s spectra analysis of LNA, LNA-w, and LNAw-650. For LNA, the highest peak (531.6 eV) was assigned to CO3and OH species, indicating high amounts of residual Li2CO3and LiOH compounds on the surface. A lower peak (529.1 eV) was assigned to lattice oxygen in the metal framework. For LNA-w, the highest peak was assigned to lattice oxygen in the metal framework, indicating that residual LiOH compounds were completely washed away because of their high solubility in water. Most of the Li2CO3compounds, which were less soluble, were also washed away. The surface analysis of LNA-w650 confirmed the titration results in Table 2; the amount of residual lithium compounds on the surface was higher for LNA-w-650 than for LNA-w. However, this amount was not as high as the amount for LNA. As shown in Figure 3d–f, the Ni 2p spectra analysis for LNA revealed a low count level due to the high amount of residual lithium compounds on the surface. After washing residual lithium compounds, the Ni 2p spectra analysis showed a higher count level for LNA-w compared to LNA. The Ni 2p3/2 peak was fitted in Ni2+(854.4 eV), indicating a delithiated phase NiO,[23] and Ni3+(856 eV) from the LNA structure.[12] Unlike LNA, LNAw’s Ni2+peak was higher than its Ni3+peak, indicating that a high amount of delithiated phase NiO formed on the surface with the reaction of H2O and LiNiO2.[22] However, after secondary heat treatment, the level of Ni2+decreased compared to Ni3+,indicating that NiO was oxidized and re-lithiated with existing Li compounds on the surface to LiNiO2.[12,22] 2.1.2. Electrochemical Performance of LNA As shown in Figure 4a, the electrochemical results from the full cell tests confirmed that an increased amount of NiO phase on the LNA-w surface decreased ionic conductivity and resulted in an initial capacity of 187.6 mAh g−1(compared to 203.1 mAh g−1 Figure 3. XPS spectra of a–c) O1s and d–f) Ni2p for pristine (LNA), washed (LNA-w), and secondary heat-treated (LNA-w-650) samples. Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (4 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Figure 4. a) Specific discharge capacities at 0.2 C for every 100 cycles until 1200 cycles, b) first cycle voltage curve, c) last cycle voltage curve, d–f) differential capacity (dQ/dV) profiles for every 200 cycles for pristine (LNA), washed (LNA-w), and secondary heat-treated (LNA-w-650) samples, respectively. Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (5 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com for LNA). However, the secondary heat treatment restored the capacity of LNA-w, increasing it to 203.9 mAh g−1.Asshownin Figure 4b, LNA-w had a higher initial charge voltage and a lower initial discharge voltage than LNA, confirming the lower conductivities of the material because of the NiO phase on the surface. However, Figure 4b also shows that LNA-w-650 and LNA had almost similar initial charge-discharge voltage profiles. This can be explained by the decrease in Ni2+on the surface due to the secondary heat treatment (Figure 3f). During the first 200 cycles of the LNA cell, the capacity decreased from 203.1 mAh g−1 rapidly to 126.9 mAh g−1. After 200 cycles, LNA-w and LNA-w650 had better capacity retention (159.4 and 162.4, respectively). After 1200 cycles, the capacity retention results for LNA, LNA-w, and LNAw-650 were 77.5, 135.8, and 131.6 mAh g−1, respectively. Full cells were tap charged at 4.2 V until the current decreased to 0.03 C. This type of charge is very harmful for LiNiO2cathode material because it increases high-volume lattice changes during a high-voltage H2 to H3 phase transitions.[24] The LiNiO2 cathode material underwent multiple phase transitions during charge and discharge from the first hexagonal (H1) phase to the monoclinic (M) phase, the M phase to the H2 phase, and the H2 phase to the H3 phase.[25,26] These phase transitions are shown as peaks in Figure 4d. Washing seemed to improve the cyclability properties of the LNA-w cell. This could be explained by the lower ionic conductivity of the surface due to the presence of delithiated phase NiO, which suppressed the H2 to H3 phase transition. Changes in H2 to H3 phase transition were confirmed by the differential capacity (dQ/dV) profiles shown in Figure 4d–f. For LNA and LNA-w-650, there were clear peaks for H2 to H3; for LNA-w, the peak almost totally disappeared. Additional dQ/dV profiles showed that, for LNA, the H2 to H3 peak completely disappeared after 200 cycles, and most of the phase transition peaks disappeared after 600 cycles. For LNA-w and LNA-w-650, H2 to H3 disappeared after 200 cycles; the other phase transition peaks remained. LNA’s poor cyclability properties could be related to a high level of charging at high voltage and an increased amount of high-volume lattice changes during the H2 to H3 phase transition and a residual lithium reaction with electrolyte and gas formation during charging. 2.2. Gas Investigations during the Cell Cycling (GC-MS) Figure 5 shows the gas measuring point in the voltage curve and the percentages of different gases produced during the first cycle charge and discharge, as measured by Gas chromatographymass spectrometry (GC-MS). During the charge, the major gas components were carbon oxides, hydrocarbons, and carbonates. Jung et al.[27] proposed that, for high-nickel cathodes, CO2/CO evolution primarily occurs because O2is released during the H2 to H3 phase transition and reacts with ethylene carbonate (EC) in the electrolyte according to Equation (1):[27,28] EC +2O2→2CO2+CO +2H2O(1) Moreover, residual lithium on the particle surface can oxidize and increase CO2production, which can explain the higher CO2 production for LNA than for LNA-w and LNA-w-650 as shown in Figure 6a.[29,30] As shown in Figures 5 and 6a, the main difference between the charge and discharge results was that, during discharge, fewer hydrocarbons were produced, and the total amount of gas was lower. The main difference between the studied samples was that cycling of LNA produced much more carbonyls, ethers, halogenated alkanes, and fluorosilanes than cycling of LNA-w and LNAw-650. Fluorosilanes indicated the amount of HF species formed when electrolyte salt lithium hexafluorophosphate (LiPF6) reacts with water (H2O) (Equation (2)) and can further attack transition metal (TM) oxide to produce slightly soluble TMF2and H2O (Equation (3)).[31] Chemical decomposition of Li2CO3with HF according to Equation (4) accelerate electrolyte decomposition reaction and increase CO2production.[27–31] LiPF6+H2O→LiF +POF3+2HF (2) TMO +2HF →TMF2+H2O(3) Li2CO3+HF →CO2+H2O+LiF (4) Fluorosilanes are detected because HF etches the protective methylsilyl layer on the GC column and forms fluorotrimethylsilane. Figure 6b shows the correlation between the fluorosilane species produced and the amount of residual lithium on the secondary particle surface, as measured by titration. The results clearly indicate that the residual lithium on the particle surface accelerated HF production. 2.3. Electrochemical Performances and Effects of the Charging Protocol To clarify the effects of high voltage charging on electrochemical performance, new cells were tested without tap charge (CC) and a formation charge voltage limited to 4.1 V. The results were compared with the cell tested with tap charge (CCCV) (Figure 7a–f). For the cells tested without tap charge—LNA*, LNA-w*, and LNA-w-650*—the initial discharges were 193.8, 170.3, and 192.3 mAh g−1, respectively. These results were ≈10 to 17 mAh g−1lower than the results of the cells tested with tap charge. The cycling properties significantly improved for LNA*, with a capacity of 155.5 mAh g−1(80.2% retention) after 1200 cycles compared to 77 mAh g−1(38% retention) for LNA. The voltage profiles of LNA and LNA* (Figure 7b) were quite similar for the first cycle. After 1200 cycles, most of the LNA charge occurred in the tap charge mode and the discharge voltage dropped very fast; however, the voltage profiles for LNA* revealed normal voltage plateaus after 1200 cycles. Despite LNA-w’s higher initial capacity, the cycling stability levels of LNA-w* and LNA-w were quite similar. After 200 cycles, the capacities were at the same level; after 1200 cycles, LNA-w* and LNA-w had capacities of 139.4 and 135.9 mAh g−1,respectively (Figure 7c). However, the discharge voltage after 1200 cycles was higher for LNA-w* than for LNA-w (Figure 7d). After 1200 cycles, LNA-w-650* had a better capacity (144.3 mAh g−1(75.0% retention)) than LNA-w-650 (132.8 mAh g−1(65.1% retention)). As shown in Figure 8a, the dQ/dV profile revealed that the H2 to H3 phase transition was more stable for LNA* than for LNA (Figure 4d). However, the H2 to H3 peak faded more than the Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (6 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com Figure 5. The gas formulation during the charge–discharge voltage curve with marked measurement point and percentages of gases (detected by GCMS) of a) LNA, b) LNA-w, and c) LNA-w-650. Small 2023, 2305349 © 2023 The Authors. Small published by Wiley-VCH GmbH 2305349 (7 of 11) 16136829, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202305349 by University Of Jyväskylä Library, Wiley Online Library on [18/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License