Scaling up Production and Analysis of Li-rich Nickel Manganese Oxide Cathode for Lithium-ion Batteries

至今，在目前所開發的鋰離子電池正極材料當中，層狀富鋰錳過度金屬氧化物這一系列的材料，xLi2MnO3•(1-x)Li(Mn, M)O2 (M= Mn, Ni, Co)，具有大於250 mAh/g的理論電容量以及平均將近4 V的氧化還原電位，因此擁有1000 Wh/kg左右的潛力理論比能量，高於其他一般正極材料如Li2MnO4，LiNi0.33Co0.33Mn0.33O2，還有 LiFePO4。 在具有如此多優點的情況下，在本篇論文中，此正極材料將在台灣立凱電能科技股份有限公司(Advanced Lithium Electrochemistry Co., Ltd.)的研發部進行試量產。由於是直接將實驗室的製程加以放大，所製備出的成品不論是物理性質還是電化學特性都需要再加以優化，在掃描式電子顯微鏡(SEM)、X光繞射(XRD)、晶粒大小分析、電化學交流阻抗分析(EIS)等工具的分析下，可以得知燒結時的條件對於成品來說是非常重要的，再藉由一步步的修改燒結時間、所使用的高溫爐、燒結溫度等，終於克服了第一圈充電時無法產生具有電化學活性的MnO2的問題，而成功地以每一批生產出100克成品的產率製備出此正極材料。此外，在安全性測試中，此正極材料之半電池具有不燃燒不爆炸的表現，展現出高度的安全性。 除了前述之優點之外，此正極材料也同時面臨一些基本的限制，像是過大的第一圈不可逆電容量、電壓不穩定、電容量衰退、較差的高速充放電性能等，在本論文接下來的章節中，分別根據在低電壓與高電壓的循環充放電測試，來探討在循環充放電中電壓衰退的主因，充放電速率為0.1 C (20 mA/g)與0.3 C (60 mA/g)。在充放電曲線圖、電容量對電壓微分圖與正規化的充放電曲線圖中，高電壓測試的電池具有較嚴重的極化現象，而低電壓測試的電池則是有著較明顯的電壓衰退。就0.1 C充放電速率而言，低電壓段產生的相轉變是造成電壓衰退的主因，而在0.3 C的充放電速率下，高低電壓段皆具有相似的電壓衰退現象，所以相轉變與內阻具有同等的貢獻。 除此之外，為了要解決此正極材料較差的高速充放電性能，本論文的最後嘗試以磷酸鋰鐵正極材料來做混合式的正極材料，希望能藉由磷酸鋰鐵平直且穩定的充放電反應平台來提升高速充放電性能。在一開始，此實驗使用軟體模擬的方式，試著先預估在各種比例下，混合後的高速充放電表現，結果指出混合式的正極材料對於快充快放的提升效果優於慢充快放。而在實際的實驗中，使用一半富鋰材料一半磷酸鋰鐵的比例，所測試出來的結果卻不如預期，只有在快充快放以3 C的速率做充放電時，實驗值會優於模擬值。推測可能是因為此兩種正極材料的粒徑分布過於懸殊，以至於鋰離子的擴散速率受到阻礙而下降，進而影響混合式正極材料的測試結果。Among the reported cathode materials so far in the lithium-ion batteries, the class of layered lithium-rich manganese-transition metal oxide composite cathode (abbreviated as LrMOs), xLi2MnO3•(1-x)Li(Mn, M)O2 (M= Mn, Ni, Co), possesses potentially high specific capacity more than 250mAh/g and high average redox potential near 4 V and therefore potential energy density nearly 1000 Wh/kg, which is much higher than those of Li2MnO4, LiNi0.33Co0.33Mn0.33O2 and LiFePO4. With so many advantages, in this thesis, this cathode material wound try to produce in a large amount in Advanced Lithium Electrochemistry Co., Ltd. (Aleees). By directly enlarging the procedure in lab, not only the physical properties but the electrochemical performance needed to be optimized. According to the analyses of Scanning Electron Microscopy (SEM), X-ray Diffraction (XRD), crystal size analysis, and electrochemical impedance spectroscopy (EIS), the calcination condition was considered to be a critical factor for its electrochemical properties. Through the modification of the calcination time, heat furnace, and calcination temperature, step by step, the main restriction for the scaling up product that the produce of electrochemical activated MnO2 in the initial charge was overcome. The perfect LrMOs product was successfully synthesized with the production rate of 100 g per batch in Aleees. Furthermore, there was no burn or explosion throughout the security tests for the scaling up LrMOs cell, demonstrated the superb safety of this cathode material. Except for the advantages, on the contrary, LrMOs encountered some fundamental limitations like large first irreversibility, voltage instability, capacity fading, and poor rate performance. Here, the main reason for the voltage fading had been investigated by the cycling test in the low and high voltage windows under 0.1 C (20 mA/g) and 0.3 C (60 mA/g). From the charge/discharge performance, the differential capacity versus voltage (dQ/dV), and normalization curves, the high voltage window cell behaved a more serious polarization and the low voltage window cell had a more apparent voltage fading because of the phase transformation. In conclusion, the low voltage window cell expressed a more prominent phenomenon of voltage fading in 0.1 C (20 mA/g), indicating that the phase transformation was the major factor for the voltage fading. As for 0.3 C (60 mA/g), both cells had similar voltage fading, showing that both phase transformation and internal impedance had similar contributions for the voltage fading. Besides, to solve the obstacle of poor rate performance, especially in the high current density, the effects of cathode mixing had been investigated, by LiFePO4, due to its flat and long plateau. In the beginning, the simulation of cathode mixing was implemented to estimate the results for the cathode mixing with different ratios between them. The results indicated that the influence for the rate performance with fast charging rate was stronger than with slow charging rate. Afterwards, compared with the simulation data, the experimental rate performance results were promulgated in the ratio of 50 : 50 for LrMOs and LiFePO4. However, the experimental rate performance test revealed an inferior results than the simulation. Only when the C-rate in 3 C (525 mA/g), with fast charging rate at different C-rates, the experimental energy density showed a higher value than the simulation one. Maybe, the relatively different particle size distribution of these two cathode materials was the main reason, which may result in the poor lithium ion diffusion between the particles