Multi-physics Field Coupling Principle Of Lithium Plating In Pouch Cells

1. Lithium Plating Phenomenon

Lithium plating is a common anomaly in lithium-ion batteries during the use of lithium-ion batteries. The direct cause of lithium plating phenomenon is that the anode material can not provide timely accommodation for free lithium ion cavities, or free lithium ions can not reach the cavities in time, leading to the accumulation of lithium ions, resulting in lithium ions to generate lithium metal reduction reaction. From the general direction of classification, lithium-ion battery lithium plating causes are divided into five categories: (1) anode aging, lithium storage residual insufficient lithium plating; (2) fast charging, overcharging, and low-temperature charging and other circumstances caused by lithium plating; (3) embedded lithium path anomalies caused by the lithium plating; (4) positive, negative pole, diaphragm, and other anomalies caused by lithium plating ; (5) special reasons for lithium plating of the fixed position.

The precipitated lithium forms lithium dendrites, which have a very strong reducing activity and a large specific surface area, and their continued growth can cause many hazards. For example, (a) excessive growth of lithium dendrites may puncture the diaphragm, leading to positive and negative short circuit, and even cause lithium-ion battery fire and explosion; (b) lithium dendrites can not be fully reversible oxidation in the discharge process, may be fractured from the root, the formation of “dead lithium”, leading to loss of active lithium inside the battery, reducing the coulometric efficiency, lower capacity; (c) lithium plating into fixed position lithium plating, which has very strong reducing activity and large specific surface area, and its continued growth will bring many hazards. (c) The newly generated lithium metal surface reacts with the electrolyte to generate a new SEI film, and at the same time consumes the electrolyte, leading to an increase in the internal resistance of the battery, local loss of activity, and a decrease in capacity.

Based on the above reasons, the causes of anode lithium plating phenomenon and its boundary conditions are hot research topics of great concern to lithium-ion battery manufacturers and application customers. Therefore, the detection and prevention of lithium plating has become a problem that IEST hopes to solve.

2. Detection Programs

Methods for detecting lithium plating can be categorized into three types:

2.1 Detection of the negative electrode potential VS Li/Li+, e.g., using a reference electrode or half-cell. Under certain operating conditions, especially when charging at low temperatures and/or high currents, lithium ions are deposited on the surface of the negative electrode instead of being embedded in the graphite lattice. This occurs only when the negative electrode potential is below 0V.

2.2 The attenuation of the battery due to lithium plating is detected by a variety of electrical signals. For example, the spectrum is analyzed by electrochemical impedance method (rise in internal resistance), abnormalities in voltage by differential voltage method, changes in elements by neutron spectrum method, and so on.

2.3 Physical/chemical analysis of lithium metal after battery disassembly.

Method (1) i.e. measurement of the negative potential in a half-cell requires a built-in reference electrode, which requires major changes to the battery structure, and the reference electrode may interfere with the electrode process and negatively affect the measurement.

Method (2) Regularly checking for degradation of cell performance, the main parameter is the decay of cell capacity, which is directly related to the loss of active lithium. However, a large portion of lithium deposition is reversible, and precipitated lithium can also be reversed by subsequent chemical intercalation into the graphite lattice, or re-dissolution upon discharge. In short, it is difficult to balance convenience, practicality and economy with these methods.

IEST provides a simple and fast method to determine the lithium plating window: A thickness expansion meter with a resolution of 0.1 µm was designed to accurately measure thickness changes during battery cycling. Under a given cycling condition, it plots the battery thickness-SOC curve when the battery is charged and discharged at constant pressure or constant gap. In the case of lithium plating, an abnormal increase in cell thickness can be clearly observed. The change in cell thickness can be used as an indicator for non-destructive lithium plating detection. The thickness indicator has higher sensitivity compared to methods based on capacity, impedance, voltage, etc., and can give rapid results.

3. Lithium Expansion Experiment

In this experiment, an in-situ swelling analyzer (SWE) is used to quantitatively evaluate the potential window for lithium plating in cells at different temperatures, providing a new method for R&D personnel to develop fast charging strategies at different temperatures. Firstly, NCM pouch cells with specific SOC are prepared, and the parameters of the test cells are shown in Table (b) in Figure 1.

Figure 1. IEST SWE instrument and battery information

Set the charge current of the battery cell as 0.5C and the discharge current as 1C, and place it in the test chamber of the in-situ swelling analyzer of the high and low temperature chamber, as shown in Figure 1(a). The batteries were charged and discharged at 25℃, 15℃, 10℃, and 0℃, and the expansion thickness change curve of the cells was monitored in real time with the in-situ swelling analyzer simultaneously (5kg constant pressure mode).

The pouch cells were fully charged at four different temperatures with the same multiplication rate, and as shown in Figure 2, the maximum thickness expansion in the four states of 25℃, 15℃, 10℃, and 0℃ were 70.6μm, 80.4μm, 80.1μm, and 92.4μm, and the expansion rates were 2.08%, 2.35%, 2.35%, and 2.71%, respectively. From Fig. 2(a), it can be seen that there are obvious differences between the 0°C curves and the 10°C, 15°C, and 25°C curves, especially the slopes of the expansion thicknesses in the high SOC state, and it is initially suspected that lithium plating occurs in the 0°C core.

The differential capacity curves at different temperatures were further analyzed, as shown in Figure 2(b). From the figure, it can be seen that the peak positions of the differential capacity curves corresponding to the four temperatures are synchronized with the sudden change in the thickness expansion rate, indicating that the thickness change of the charging process of this core is due to the phase change of de-embedded lithium in the positive and negative electrode materials. And as the temperature decreases, the phase transition peak is moving to the right, indicating that the polarization is increasing. Comparing the thickness expansion curves at different temperatures, in the thickness curve corresponding to 0°C, the thickness expansion corresponding to the position with a sudden change in slope is obviously larger than that at the other three temperatures, and the thickness expansion curve is differentiated, as shown in Figure 2(c). It is also obvious from the figure that the expansion thickness per unit capacity of the 0°C core is also slightly larger than that of the other three groups after the charging capacity is greater than 30% SOC and the voltage reaches 4.27 V. This is most likely due to the fact that with the decrease of the test temperature, the polarization of the core increases and there is lithium plating on the surface of the negative electrode, which leads to a larger slope of the thickness expansion.

Figure 2. (a) Charge capacity-SOC-thickness curves of pouch cells at different temperatures

Figure 2. (b) Differential capacity-voltage-thickness curves of pouch cell at different temperatures

Figure 2. (c) Differential thickness-SOC-thickness curves of flexible pack cells at different temperatures

3. Dendrite Growth Analysis

Through the expansion lithium plating experiments, the thickness variation data of the cell cycling at different temperatures were obtained, in which the thickness capacity differential curves indicate the time, capacity and temperature windows corresponding to lithium plating. Then, through further theoretical analysis, an attempt was made to calculate the specific amount of lithium plating under this condition. Battery expansion at different temperatures is a complex multi-physics field coupling problem involving ion concentration field, electric potential field, physical phase field and temperature field. We theoretically analyze the morphology and conditions of dendrite formation by building a continuous thermodynamic phase field model. In the model, three nonlinear equations are usually included:

3.1 Phase field equation

The phase transition of elemental lithium from the ionic to the metallic state with time is proportional to the change in the free energy of the system. Where the free energy of the system can contain the energy generated by the chemical reaction of the battery, the energy of the elastic-plastic change of the material, the thermal energy and so on.

3.2 Ion diffusion equation

i.e., Fick’s second law, the change in solid-phase lithium ion concentration with time is related to the ion concentration gradient along the radius inside the particle. The significance of this equation is that the concentration change in solid phase ion concentration at any given moment is related to its diffusion coefficient and concentration gradient.

3.3 Potential distribution equation

For the electrostatic potential distribution, assuming charge neutrality in the system, the current density is considered to be conserved by Poisson’s equation, including a source term to represent the charges entering or leaving due to electrochemical reactions.

By entering the set of nonlinear equations (a)(b)(c) into Comsol software, the dendrite morphology can be calculated for a specific temperature, multiplicity and voltage. This is shown in Figure 3.

Figure 3. Lithium dendrite simulation results

4. Comparative Analysis Of The Volume Change Of Lithium Plating And Graphite Embedded Lithium

In graphite/ternary batteries, the volume change of anode is very small (≤1%), while the volume change of graphite full embedded lithium state is 10%. Therefore, the volume change of the battery is mainly determined by the graphite anode. Of all the carbon atoms in the graphite material, one ion can be stored for every six carbon atoms, so the maximum amount of lithium ions that can be stored in graphite is one lithium atom for every six carbon atoms, and the minimum amount of lithium ions that can be stored is zero. Therefore, when talking about the degree of lithiation of a graphite electrode, we use the symbol LixC6, where 0 ≤ x ≤ 1. Clearly, at the atomic level, for any C6 ring, either only one lithium atom is inserted in the ring or no lithium atoms are inserted. However, when considering the entire electrode, only a fraction of the C6 rings will have lithium ions embedded in them, and this fraction is the value of x. When the battery is charged, the negative electrode is highly lithiated, and in the lithium-rich state the x in LixC6 is close to 1. When the battery is discharged, there is a large amount of de-embedded lithium atoms in the negative electrode, and in the lithium-poor state the x in LixC6 is close to 0. The lithium-rich state is the same as the lithium-poor state.

The electrochemical insertion reaction of desolvated lithium ions in graphite occurs in the potential range of 0 to 0.25 V vs Li/Li+. The insertion reaction occurs at well-defined voltage plateaus, and there are also well-defined insertion compounds at the beginning and end of the voltage plateaus. Experiments show that the lithium intercalation step is measurable. Hexagonal (ABABAB) and rhombic (ABCABC) graphite structures are transformed into an AAAAAAA stacking sequence with inserted lithium during the lithium embedding process, respectively. The lithium is located in the center of the C6 ring between the two graphene layers. Thus, the capacity of graphite depends on the number of available graphene layers. If a very well structured graphite (e.g. natural graphite) is used, it can be almost fully charged up to a theoretical capacity of 372 mAh/g in a multiplicative charging scenario, and Figure 4 shows the anode changes during the lithium insertion process.

Figure 4. Graphite-embedded lithium potential curves and schematic diagrams

The insertion of lithium ions into the anode lattice fills the free gap, so that the volume of the intercalation compound is less than the sum of the volumes of the two individual materials. The increased thickness of the intercalation material does not cause significant expansion of the battery. However, lithium metal deposited on the surface of the negative electrode significantly increases the thickness of the soft pack battery. Transferring a charge of 1Ah from the cathode to the anode, the amount of lithium material is 37.31 mmol. every six carbon atoms provide space for one lithium atom to insert. Therefore, for a charge of 1Ah, 0.2239 mol of carbon is consumed. The volume of this portion of carbon is 1.189 cm³. According to the literature, it is known that the carbon-based anode material expands by about 10% due to the insertion of lithium ions. Therefore, charging 1Ah results in a volume change of 0.12cm³. When lithium ions are reduced to lithium metal precipitated on the surface of the anode, the volume of lithium metal reduced by 1Ah charge is 0.49cm³, assuming dense deposition of lithium. If the lithium forms a dendritic morphology (Figure 3), the precipitated lithium becomes fluffy and the volume increases even more. Assuming that the percentage of lithium dendrites is x%, combined with the morphology of lithium dendrites under this condition, and based on the data measured by the in-situ expansion equipment at the specified constant pressure and the measured Young’s modulus of the cell, the calculations lead to a lithium precipitation of 4% under the cycling condition of 0°0.5C in Figure 2(b).

5. Summary

The in-situ swelling analyzer developed by IEST can accurately detect the thickness change of batteries under specified working condition cycles. The lithium plating temperature and SOC window under specific conditions can be accurately determined by thickness differentiation and other algorithms, and the amount of lithium ions consumed by lithium dendrites can be roughly calculated by combining with lithium dendrite morphology simulated by comsol.

6. References

[1] B. Bitzer, Gruhle. A new method for detecting lithium plating by measuring the cell thickness. Journal of Power Sources, 262 (2014) 297~302.

[2] Modulation of dendritic patterns during electrodeposition: A nonlinear phase-field model