It can be seen from the electrode potential that the positive electrode material of lithium ion battery has strong oxidizing property in the charged state, and the charged positive electrode material may oxidize or self-decompose the solvent, and these properties are closely related to the safety performance of the battery.
The self-decomposition reaction of LiCoO2 and LiNiO2 can be measured by thermal analysis and X-ray diffraction. The thermal decomposition reaction of electrochemical delithiation material Li1-yCoO2 (0.40.7, Li1-yNiO2 forms LixNi1-xO of rock-salt structure, and when y<0.7, Li1-yNiO2 is layered structure. When y=0.5, the reaction also follows the formula. However, it is difficult to determine whether the spinel phase is formed because the XRD patterns of the layered and spinel structures are similar. The spinel phase is easy to form in nickel-rich compounds, but thermal analysis shows that LiNi2O4 is difficult to form in a relatively short time during the conversion of Li0.5NiO2 to LiNi2O4, and it takes dozens of hours in high-purity LiNiO2 . The difference between the two reactions in Li0.5NiO2 is oxygen evolution, which can be determined by thermogravimetric analysis.
Li0.5CoO2=(1/2)LiCoO2+(1/6)Co3O4+(1/6)O2
Li0.5NiO2=(1/2LiNi2O4)=(3/2)Li0.33Ni0.67O+(1/4)O2
The main difference between the above two equations is that the amount of oxygen evolution is different. Since the O2 produced by Li0.5NiO2 is 1.5 times that of Li0.5 CoO2, LiNiO2 is easier to generate heat than LiCoO2.
In the absence of electrolyte, the thermal analysis of the product after charging only shows the information of self-decomposition, so it is possible to know the thermal performance of the decomposition reaction, the release temperature and amount of oxygen, etc. However, the endothermic reaction between the electrolyte and oxygen can lead to thermal runaway of the battery in the charged state, therefore, the safety of the battery should be evaluated in the presence of the electrolyte.
Dahn et al. evaluated the heat generated by cathode materials in various states of charge in the presence of electrolytes by DSC methods. Figure 1 (a) and (b) depict the charge capacity versus heat generation rate. For all cathode materials, the heat generation rate increases as the charge capacity increases. In the figure, the 4V cathode materials such as LiCoO2, LiMn2O4 and LiNi0.8Co0.2O2 are almost linear, but LiNi3/8 Co1/4 Mn3/8 O2 deviates greatly. The 3V LiFePO4 has a lower heat yield, which is due to its weak oxidation or the covalent bond stability of PO43-, while the performance of LiNi3/8 Co1/4 Mn3/8O2 is strange, and its heat yield is very low.
These data show that LiNi0.5Mn0.5O2based cathode materials have excellent thermal stability, such as LiNi1/3Co1/3Mn1/3O2.
LiNiO2 exhibits a unique heat production rate, which suddenly increases when Li0.3NiO2 containing the NiO2 phase is formed. Therefore, if the generation of the NiO2 phase can be prevented, the sudden increase in the heat generation rate can be suppressed. In fact, the heat production rate can be greatly reduced by doping with 20% cobalt. Many studies have shown that doping cobalt in LiNiO2 can significantly improve its thermal stability, and some people think that the safety of LiNiO2 can be greatly improved by replacing 1/4Ni with Al. The electrochemically active material in the charged state reacts with the electrolyte at the two-phase interface, and the reduction of the particle size of the electrochemically active material will lead to an increase in the heat generation rate. Therefore, the use of nanoscale materials to improve the rate capability will increase the safety risk.
The safety performance of polymer battery cathode materials has also been evaluated, Li0.6CoO2, Li0.6NiO2 and Li0.23Mn2O4 have about several W/g exothermic peaks in the range of 200~300 °C; organic materials and cathode materials are thermally decomposed The reaction between the generated oxygen is almost inevitable. The evaluation of the safety performance of lithium-ion batteries is detailed in the paper by Tobishima et al.
