Summary of research on cycle life of lithium batteries

Lithium ion batteries are widely used in various fields, such as electronic products, power tools, electric vehicles, 

and energy storage, due to their high energy density, no memory effect, low self discharge, and long cycle life. 

The overall performance of batteries can be divided into two categories: electrical performance and reliability, 

and lifespan is one of the important indicators to measure their electrical performance.

For energy type batteries, it is generally believed that the end of their lifespan occurs when the available capacity of the battery decays to 80% of its initial capacity.

 The lifespan of a battery includes cycle life and calendar life. 

The former refers to the number of cycles a battery undergoes under a certain charging and discharging schedule until the end of its lifespan, 

while the latter refers to the time required for the battery to be stored in a certain state until the end of its lifespan.

Lithium batteries undergo many complex physical and chemical reactions during the charging and discharging process, 

so there are many factors that affect the cycle life of lithium batteries. On the other hand, cycle life testing is often time-consuming and costly,

 and the correct evaluation of battery life has a certain guiding role in the production and development of lithium batteries and battery health management systems.

1、Factors affecting cycle life

①The aging and decline of battery materials

The materials inside lithium batteries mainly include: positive and negative electrode active materials, binders, conductive agents, current collectors, separators, and electrolytes. 

During the use of lithium batteries, these materials will experience a certain degree of degradation and aging. 

Tang Zhiyuan and others believed that the capacity attenuation factors of lithium manganate battery include: 

dissolution of cathode material, phase change of electrode material, decomposition of electrolyte, formation of interface facial mask and collector corrosion.

Vetter et al. conducted a systematic and in-depth analysis of the mechanisms of changes in the positive, negative, and electrolyte of batteries during cycling.

 The author believes that the formation and subsequent growth of the negative electrode SEI film will be accompanied by irreversible loss of active lithium,

 and the SEI film does not have true solid electrolyte function.

 In addition to lithium ions, the diffusion and migration of other substances can lead to gas generation and particle rupture.

 In addition, the change in material volume and the precipitation of metallic lithium during the cycling process can also lead to capacity loss. 

Aurbach et al. disassembled the positive and negative electrode plates of lithium cobalt oxide batteries after cycling at temperatures of 25 and 40 ℃. SEM, XRD,

 and FTIR test results showed that both positive and negative electrode active materials were lost. 

Li Yang et al. analyzed the electrical performance of a lithium iron phosphate power battery with 6000 cycles, and found that its capacity retention rate was 84.87%. 

The AC internal resistance increased by 18.25%, and the DC internal resistance increased by 66%. 

The author disassembled the cycled battery and conducted performance tests and SEM analysis on the button type battery. It was found that the performance of the negative electrode material deteriorated rapidly after cycling, 

and it was believed that the expansion of the negative electrode volume and the thickening of the SEI film were the main influencing factors.

② Charging and discharging system

The charging and discharging system mainly includes three aspects: charging and discharging methods, rate, and cut-off conditions.

 In terms of charging methods, American scientist Mas once proposed the concept of the optimal charging curve, 

believing that the optimal charging current of a battery gradually decreases with the extension of charging time: I=I0e - α t. In the formula: I is the acceptable charging current;

 I0 is the maximum initial current at time t=0; T is the charging time; α is the decay constant. 

Below the curve is the rechargeable area, where charging will not cause damage to the battery.

 If the charging current exceeds this area, polarization will intensify, 

which not only fails to improve charging efficiency but also leads to severe gas evolution and shortens battery life. 

At present, most of the research on charging methods is based on the Mas theory, 

which aims to make the charging current as close to the curve as possible.

He Qiusheng and others conducted a comprehensive comparison of several common charging methods and found that constant current charging, 

due to excessive current in the later stage, causes internal gas evolution and damage to the battery; 

Constant voltage charging, on the other hand, causes excessive current during the initial charging stage, which directly damages the battery; 

Constant current and constant voltage charging, as well as the step constant current charging method, 

overcome the shortcomings of constant current and constant voltage charging and are currently widely used;

 Reverse pulse charging can effectively eliminate polarization, but it has a certain impact on lifespan.

The charge discharge rate and cut-off conditions also have a significant impact on the battery cycle life.

 Li Yan et al. studied the cycle performance of 18650 lithium cobalt oxide battery under different discharge rates, 

and found that the capacity loss rates after 300 cycles at 0.5C, 1C and 2C discharge rates were 10.5%, 14.2% and 18.8% respectively.

 Through analysis,

 it was found that the change in the structure of the positive material and the thickening of the negative surface facial mask would lead to the reduction of the number of lithium ions 

and the blocking of diffusion channels, 

which would lead to the decline of the battery capacity.

K. Maher et al. raised the charging cut-off voltage of lithium cobalt oxide batteries from 4.2V to 4.9V,

 and found that the structure of the electrode material had changed by testing the entropy change curves of the electrodes at different SOC after charging.

③Temperature

Different types of lithium batteries have different optimal operating temperatures, and temperatures that are too high or too low can have an impact on the battery's lifespan. 

Ramadass et al. reported the effect of temperature on the cycling performance of Sony 18650 lithium cobalt oxide batteries. 

The study found that when the test temperature exceeded 50 ℃, the battery decay was significantly faster than at normal temperature and 45 ℃ (Figure 3). 

The capacity decay at high temperatures was attributed to the decomposition and regeneration of the battery's negative SEI film, 

the loss of active lithium, and the increase in negative electrode impedance.

Song Haishen et al. compared the electrical performance of 18650 lithium iron phosphate/graphite power batteries at different temperatures and obtained similar results:

 under normal temperature cycling, the capacity decay of the battery is relatively slow,

 while under high temperature conditions of 55 and 65 ℃, the battery exhibits rapid failure behavior. 

The author believes that the trace iron deposited on the graphite anode will catalyze the formation of its interfacial facial mask, which has a certain influence on the capacity attenuation.

Zhang et al. studied the performance of lithium batteries at low temperatures and found that the capacity of the battery sharply decreases when the temperature is below -10 ℃.

 They analyzed that the poor low-temperature performance is not only due to the decrease in ion conductivity of the electrolyte, but also related to the electrode material.

 The author compared the EIS curves of the full cell and the symmetrical electrodes with temperature and found that when the temperature was below -10 ℃, 

the impedance of both the full cell and half cell showed an upward trend, especially the charge transfer impedance, which suddenly increased and dominated.

④ Monomer consistency

Battery packs typically consist of hundreds or thousands of individual cells connected in series or parallel. 

In addition to the aforementioned influencing factors, cell consistency is another important factor in their cycle life. 

Due to differences in materials and manufacturing processes, it is difficult to ensure the consistency of individual lithium batteries. 

In terms of materials, the uniformity of positive and negative electrode materials and electrolyte is important,

 and lithium batteries produced from the same material and batch often have relatively good consistency. 

In terms of manufacturing, the production process of lithium batteries is very complex, and each step involves multiple process parameters. 

If not controlled properly, it can lead to inconsistencies in parameters such as voltage, capacity, and internal resistance of the battery.

Wang Zhenpo et al. studied the impact of individual inconsistency on the service life of battery packs. 

They believe that the service life of a battery pack is always shorter than that of the shortest individual battery. 

A single battery with a service life of 1000 times has a service life of less than 200 times after grouping, 

and the increase in battery pack life is not proportional to the increase in battery pack life (Table 1).

Chen Qiang et al. investigated the impact of Ohmic resistance, capacity,

 and polarization differences of individual cells on the performance of series connected battery packs based on Thevenin equivalent circuits, 

and found that capacity differences had the greatest effect.

Before actual group application, batteries undergo a screening and grouping process to eliminate individual cells with significant differences in performance parameters,

 minimizing the impact of differences in battery manufacturing on usage performance. 

Batteries are generally assembled based on parameters such as capacity, voltage, internal resistance, and self discharge.

 However, rapid detection of self discharge in batteries is a research difficulty.

 The self discharge of individual batteries can lead to inconsistent SOC of each battery in the battery pack, affecting the capacity of the entire battery pack. Generally speaking, 

the higher the temperature, the greater the self discharge of the battery. If the design of the battery pack casing is not reasonable,

 the internal resistance and self discharge degree of batteries in different positions will be affected to some extent due to differences in heat dissipation.

2、Cycle life prediction

Due to the long testing time and high cost of battery cycle life, 

the establishment of life models and the evaluation and prediction of life have become research hotspots for scholars at home and abroad. 

The life prediction methods for lithium batteries can be divided into three categories based on information sources: prediction based on capacity degradation mechanism, 

prediction based on characteristic parameters, and data-driven prediction.

① Prediction based on capacity degradation mechanism

Mechanism based prediction is to infer the lifespan of a battery based on the aging and degradation mechanisms of its internal structure and materials during cycling.

 This method requires the use of basic models to describe the physical and chemical reaction processes that occur inside the battery, 

such as Ohm's law, electrochemical polarization, concentration polarization, and internal diffusion of electrode materials.

Based on the loss of active lithium during the battery cycle, Ning et al. used the first principle to simulate the capacity decline model of lithium cobalt oxide battery.

 The influencing parameters include exchange current density, DOD, interface facial mask impedance, and charging cut-off voltage. 

The author will compare the life prediction model with the measured data and find that the model is very close to the actual detection results.

Virkar proposed a degradation model for batteries based on non-equilibrium thermodynamics, considering the effects of chemical potential and SEI film on capacity degradation.

 He pointed out that there may be unbalanced cells in the series battery pack, and SEI film may also form at the interface between the positive electrode and electrolyte, leading to increased capacity degradation.

② Prediction based on feature parameters

Feature parameter based prediction refers to the use of changes in certain characteristic factors during the aging process of batteries to predict battery life.

 Currently, researchers are most concerned about the relationship between EIS and cycle life. 

Li et al. studied the change of impedance spectrum of commercial lithium cobalt oxide battery during the 1C charge discharge cycle, 

and observed the change of electrode materials by XRD, TEM and SEM. 

It was found that in the Nyquist curves of the positive and negative electrodes of the lithium battery, 

the size of the semicircle in the low frequency region corresponding to the impedance of the interface facial mask increased with the increase of the number of cycles, 

from which the cycle life of the battery could be inferred.

EIS can provide a more detailed description of battery impedance, 

but testing instruments are susceptible to external interference and difficult to effectively analyze complex spectra. Relatively speaking,

 the measurement of pulse impedance is simple and easy to implement, and can be quickly monitored online.

③ Data driven prediction

The data-driven approach refers to directly analyzing test data to uncover patterns without considering the physical and chemical reactions and mechanisms within the battery.

 It is an empirical simulation method. Common methods include time series models (AR), artificial neural network models (ANN), and correlation vector methods (RVM).

The AR model infers the predicted value of the current state based on data measured at certain previous time points, and has linear characteristics. 

Considering the nonlinear relationship between battery capacity degradation and cycle times, Luo Yue proposed an improved nonlinear AR model,

 which introduces an accelerated degradation factor in the later stage of prediction to improve the accuracy of prediction.

The ANN model is an artificial intelligence network system composed of multiple neurons according to certain rules, and is a typical nonlinear model. 

The RVM model belongs to the data regression analysis method,

 which can flexibly control overfitting and underfitting by adjusting parameters, and has the characteristics of probabilistic prediction. 

The prediction method based on internal mechanisms has better theoretical support and accuracy, but it is more complex. 

The advantage of data-driven methods is simplicity and practicality, but due to the inability of the obtained data to cover all parameters, it also has certain limitations.

3、summary

This article mainly introduces the research on the influencing factors of cycle life and life prediction models of power lithium-ion batteries. 

It can be seen that there are many factors that affect the cycle life of power lithium batteries, and the influencing factors are also different for lithium batteries with different materials and structures.

From the analysis in the article,

 it can be seen that we can extend the battery life by controlling parameters, 

such as allowing the battery to operate under appropriate temperature, rate, and charge discharge conditions. 

Relatively speaking, the factors affecting the cycle life of a battery pack are more complex, 

as these factors interact with each other and the issue of cell consistency can lead to the underutilization of the battery pack's performance, severely shortening its cycle life.

When predicting the cycle life of a battery, the establishment of an accurate, reasonable, 

and simple operational model based on the internal mechanism of the battery, a certain characteristic parameter,

 or a large amount of measured data is of great significance for the accurate evaluation of battery cycle life and further optimization of performance.