An Experimental and Modeling Study on Commercial Lithium Titanate Batteries with Different Cathode Materials

Abstract

This study presents a comparative analysis of the performance and modeling differences among lithium titanate oxide (LTO) batteries with three different cathode materials. An evaluation was conducted by performing performance tests over −20 °C to 25 °C at various current rates. Differences in open-circuit voltage curves, as well as charge and discharge capacities under different temperatures and C-rates, were systematically compared. At 25 °C, the NCM cathode enabled superior rate capability, retaining over 90% of its capacity at 8 C discharge, whereas the LCO-based cells exhibited significant capacity fade. Conversely, at −20 °C, the LCO cathode demonstrated better low-temperature performance, delivering almost 80% of its room-temperature capacity at 4 C, compared to less than 5% for the NCM cathode. The batteries were modeled using a second-order equivalent circuit model, and variations in model parameters were analyzed from the perspectives of internal resistance and electrode kinetics. The second-order equivalent circuit model revealed that the NCM-based cells had lower ohmic resistance and faster electrode kinetics. By correlating battery performance with cathode materials, this study evaluates the suitability of LTO batteries with different cathodes for various application scenarios, providing valuable insights for battery application and management.

1. Introduction

Recently, electrochemical energy storage (EES) devices are gaining increasing importance due to their favorable lifecycle, high energy density, reliability, and stability [1,2,3]. Among these, rechargeable batteries represent one of the most useful and popular EES technologies, offering superior overall performance in commercial applications such as portable electronics and electric vehicles for transportation and communication [4,5,6]. Lithium-ion batteries (LIBs) are widely regarded as having the best combination of properties among various rechargeable battery types; however, their application is often constrained by the physical limitations of electrode materials [7,8].

In particular, Lithium-Titanate-Oxide (LTO) batteries, which employ spinel-structured lithium titanate as the anode material, exhibit outstanding performance characteristics, including excellent rate capability and chemical stability [9]. As a result, LTO batteries have emerged as a leading candidate for fast-charging and automotive assistance applications [10,11,12]. In the field of new energy locomotives for rail transit, due to regional differences in usage (e.g., northeastern China requires lithium-ion batteries with superior low-temperature performance, whereas southern China imposes less stringent low-temperature requirements), comparative research on the cathode materials of lithium titanate batteries is essential during battery selection and application to better suit specific operational scenarios. Compared to lithium iron phosphate batteries and NCM batteries with graphite anodes, LTO batteries demonstrate significant advantages in start-up and regenerative braking applications. This is primarily attributed to their outstanding power density and wide operating temperature range [13,14].

However, battery performance is not determined by the negative electrode alone; the positive and negative materials work together to influence overall performance. Most existing studies focus on comparisons between graphite-anode batteries or between graphite and LTO-based cells [15,16]. For instance, Nikolian et al. investigated differences in equivalent circuit models among Nickel–Cobalt–Manganese (NCM), Lithium–Iron–Phosphate (LFP), and LTO batteries under varying temperatures [17]. Sharma et al. examined variations in electrochemical models across different cathode-anode pairs [18]. Heydarzadeh et al. analyzed discrepancies in equivalent circuit models derived using different modeling approaches for LTO, LFP, NCM, and Nickel–Cobalt–Aluminum (NCA) cells [19]. Dubarry et al. developed an automated state-of-health estimation method and compared aging models for LFP/Graphite and NCM/LTO batteries [20]. Despite these efforts, a comparative analysis of LTO batteries with different cathode materials remains lacking.

In the design and application of LTO batteries, various cathode materials such as LFP, Lithium-Cobalt-Oxide (LCO), and NCM have been adopted [21,22,23,24]. However, due to the relatively low lithium potential of LFP materials, LFP/LTO configurations often suffer from insufficient energy density, which has limited their commercial adoption. Instead, LCO or NCM are more commonly used as cathodes in commercial LTO batteries. Therefore, this study focuses on three LTO batteries with different cathode materials, providing a comparative analysis of their electrochemical performance and modeling characteristics to evaluate the impact of cathode chemistry.

The remainder of this paper is organized as follows: Section 2 describes the experimental setup. Section 3 outlines the equivalent circuit model (ECM). Section 4 presents and discusses the comparative results of the performance and models for the three battery types. Finally, Section 5 provides the concluding remarks.

2. Experimental Setup

Three lithium-ion batteries with different cathode materials but all featuring LTO as the anode material were selected. The specific cell parameters are presented in

Table 1. Batteries information.

	Cell A	Cell B	Cell C
Cathode material	NCM (LiNi0.5Co0.2Mn0.3O2)	NCM + LCO (3:7)	LCO
Anode material	LTO	LTO	LTO
Capacity	20 Ah	10 Ah	25 Ah
Nominal voltage	2.2 V	2.3 V	2.3 V
Packaging Type	Prismatic	Pouch	Pouch
Manufacturer	TOSHIBA (Tokyo, Japan)	Microvast (Huzhou, China)	RiseSun MGL (Beijing, China)

. The open-circuit voltage (OCV) of a battery reflects its electrochemical performance; however, since actual OCV testing can only be measured at a few discrete state of charge (SOC) points, the quasi-open-circuit voltage (OCV-SOC) method is generally employed to obtain complete OCV data. The OCV-SOC test involves very slow charging and discharging at a current rate (C-rate) of 0.05, followed by a 1 h rest period after each 5% SOC to capture the change in battery voltage over time.

Furthermore, to evaluate the performance of the batteries under various charge and discharge rates, a series of constant-current charge/discharge tests were conducted. Each battery was first charged or discharged at a 1 C rate until the cutoff voltage was reached. Subsequently, reverse charge and discharge cycles were performed at six different C-rates (0.5 C, 1 C, 2 C, 4 C, 6 C, and 8 C). This experimental protocol allows for the assessment of battery performance as a function of applied current rate. All tests were repeated at −20 °C, 0 °C, and 25 °C to evaluate the adaptability of the batteries under different temperature conditions. Activated each new battery by performing 10 full-charge–discharge cycles, followed by capacity cycling tests at multiple rates, with two capacity tests at each rate. Since these were commercially available lithium-ion batteries, these cycles did not cause lifespan degradation. Therefore, the capacity changes are primarily attributable to the effects of current rate and temperature, rather than cycle aging.

Finally, pulse current tests were conducted at every 5% SOC interval to parameterize the equivalent circuit model. Each pulse had a magnitude of 1 C and a duration of 60 s as shown in Figure 1.

The charge–discharge test system used was a multi-channel 5 V–100 A instrument manufactured by Arbin (Arbin Instruments, College Station, TX, USA), with a voltage accuracy of ±0.02% and a current accuracy of ±0.05%. The temperature-controlled chamber, supplied by Giant Force Co. (Beijing, China)., has an operational range of −60 °C to 100 °C and a temperature resolution of 0.1 °C.

Since only one battery of each type was used for testing, the results are limited. However, the primary objective of this study is to conduct a comparative analysis of the fundamental performance trends and modeling characteristics inherent to these three cathode chemical systems when paired with LTO anodes. Therefore, only one cell of each type is sufficient to demonstrate their capacity decay rates and kinetic differences at low temperatures.

3. Related Work on Modeling

As outlined in Section 1, the primary objective of this study is to investigate the differences among LTO batteries with different cathodes and their applicable scenarios, thus necessitating model-based analysis. A second-order ECM is adopted to simulate the battery voltage in response to the current drawn from the cell, as illustrated in Figure 2.

The model incorporates key variables including the battery OCV, ohmic resistance R₀, polarization resistances R₁ and R₂, and polarization capacitances C₁ and C₂. These parameters are influenced by battery-specific characteristics such as electrode materials and SOC. The terminal voltage U(t) is computed using Equation (1), where I(t) represents the load current.

$$\begin{array}{c}U\left(t\right)=OCV+R_{0}I\left(t\right)-v_1-v_2\end{array}$$

(1)

The dynamics of the parallel RC network voltages (v₁ and v₂) are described by Equation (2).

$$\left[\begin{array}{c}\dot{v_1}\\ \dot{v_2}\end{array}\right]= \left[\begin{array}{cc}{\displaystyle \frac{-1}{R_1{C}_1}}& 0\\ 0& {\displaystyle \frac{-1}{R_2{C}_2}}\end{array}\right]\cdot \left[\begin{array}{c}{v}_1\\ v_2\end{array}\right]-\left[\begin{array}{c}{\displaystyle \frac{1}{C_1}}\\ {\displaystyle \frac{1}{C_2}}\end{array}\right]\cdot I\left(t\right)$$

(2)

The voltages v₁ and v₂ across the two parallel RC networks are defined with initial states that depend on the operational history of the battery.

The ohmic resistance (R₀) is primarily associated with the ionic conductivity of the electrolyte and electronic resistance of the electrodes and contacts. Its significant increase at low temperatures (implied by the poor low-temperature performance) is attributed to the reduced ionic conductivity of the electrolyte.

The polarization resistances (R₁, R₂) and their associated time constants (τ₁, τ₂) are related to kinetic limitations such as charge transfer resistance at the electrode-electrolyte interfaces and solid-state diffusion within the active materials.

The model parameters (R₀, R₁, C₁, R₂, C₂) were identified by applying a nonlinear least squares fitting algorithm to the voltage response data from the pulse tests (60 s 1 C pulses). This algorithm optimizes the parameter values by minimizing the error between the model-predicted voltage and the measured voltage [25].

4. Results & Discussions

4.1. Electrochemical Performance

As can be seen from Figure 3, the OCV curve of Cell C exhibits a distinct shape compared to those of Cells A and B. Specifically, it shows a broad voltage plateau below 50% SOC, whereas Cells A and B only display minor plateaus around 70% SOC and 50% SOC, respectively.

To further compare and analyze the differences in the OCV curves of the three cells, the corresponding incremental capacity (IC) curves and differential voltage (DV) curves are also provided in Figure 4.

Figure 4a presents the IC curve of Cell A, in which two distinct peaks can be clearly observed, corresponding to phase transition processes in the NCM material. In comparison, the main peak of Cell B’s IC curve is similar in intensity to that of Cell A. However, the peak around 2.2 V is significantly weaker in Cell B in Figure 4c. This is because the main peaks of the IC curves for both NCM and LCO materials are located near 2.35 V, and the cathode of Cell B is a blend of NCM and LCO materials. As a result, the main peak reflects a combined contribution from both materials, while the peak around 2.2 V corresponds solely to the phase transition of NCM. Since NCM constitutes only a portion of the total cathode material in Cell B, this peak appears diminished.

Notably, two small peaks are observed in the IC curve of Cell B near 2.5 V and 2.6 V, which are associated with phase transitions in the monoclinic phase of LCO. These features are more clearly visible at charging capacities of 9.5 Ah and 10.5 Ah in Figure 4d. In contrast to Cells A and B, the IC curve of Cell C exhibits only one dominant peak in Figure 4e, corresponding to the extended voltage plateau observed in its OCV curve.

The charge and discharge capacities of the batteries under different temperature conditions are compared in Figure 5, Figure 6 and Figure 7. Figure 5 displays the performance at room temperature (25 °C). It can be observed that although Cell A exhibits the lowest charging capacity at low C-rates among the three cells, it delivers the highest discharge capacity. Notably, Cell B shows the smallest charging capacity at an 8 C rate, which may be attributed to its LCO cathode material. In contrast, Cell A, which uses an NCM cathode, demonstrates relatively stable capacity even at higher C-rates. This suggests that LCO-based cathodes exhibit inferior rate performance compared to NCM-based materials.

Figure 6 and Figure 7 present the charge and discharge capacities of the cells at 0 °C and −20 °C, respectively. The results clearly indicate that Cell A performs the poorest under low-temperature conditions, followed by Cell B, while Cell C exhibits the best performance. This implies that LCO-based materials possess better low-temperature performance compared to NCM material. Furthermore, blending the two cathode materials can further enhance the low-temperature capacity. It is noteworthy that at −20 °C, the battery capacities become almost negligible at C-rates above 4 C, indicating that the cells are practically unusable under such conditions. Therefore, such operating scenarios should be avoided in practical applications.

4.2. Equivalent Circuit Model Parameters

Figure 8a–c, respectively, show the comparison between the simulated voltage values obtained using the equivalent circuit model and the experimental data, along with the relative voltage error. In battery equivalent circuit models, the R₀ parameter represents the ohmic resistance of the cell and serves as a key indicator of battery performance. The ohmic resistances of three different cell types at various SOC are presented in Figure 9. It can be clearly observed that Cell B, which uses LCO as the cathode material, exhibits the highest ohmic resistance, with its resistance value gradually decreasing as SOC increases. In contrast, the ohmic resistances of Cell A and Cell C remain relatively low throughout the entire SOC range, measuring much smaller than those of Cell B.

For a more precise comparison, it is recommended to analyze the time constants of the subcircuits. These time constants reflect two distinct dynamics, corresponding to the slow and fast dynamics of battery behavior. The specific method for calculating the time constants of the two RC parallel branches is obtained from Equation (3).

$$\left\{\begin{array}{c}{\tau }_1=R_1{C}_1\\ {\tau }_2=R_2{C}_2\end{array}\right.$$

(3)

Figure 10 illustrates the time constants for each cell at different SOC levels, computed using the identified parameters of the equivalent circuit model. The time constants of Cell B vary considerably across SOC, whereas those of Cell A and Cell C remain relatively stable over the entire SOC range. Furthermore, the τ₁ value of Cell B is larger than those of Cell A and Cell C, indicating slower electrode kinetics in Cell B. This observation further supports the previously mentioned inferior rate capability of LCO material compared to NCM material. Although a larger time constant may indicate different kinetic characteristics, the cycling performance of the hybrid electrode (Battery B) could be advantageous [26]. This aspect of the research requires further investigation in the next phase.

5. Conclusions

In this study, the differences between LTO batteries with different cathode materials were compared in detail from the perspectives of both electrochemical performance and ECM parameters. The results indicate that LTO cells with an LCO cathode exhibit higher operating voltage, corresponding to greater energy density. It also has a greater usable capacity at low temperatures. In contrast, LTO cells with an NCM cathode demonstrate superior rate capability, delivering more capacity under high-current discharge conditions. Parameter identification of the equivalent circuit model further supports these findings: batteries with LCO cathodes show higher resistance, while those with NCM cathodes exhibit faster electrode kinetics. Therefore, LTO batteries with LCO cathodes are more suitable for low-temperature applications, whereas those with NCM cathodes are better suited for high-power scenarios. It should be noted that for hybrid material electrodes, their kinetic and cycling performance exhibits more complex characteristics due to potential synergistic effects between materials, necessitating further research.

From the perspective of applying new energy traction vehicles in rail transit, EIS testing is not essential when selecting battery types, as the focus is primarily on the charge/discharge capacity at different temperatures and simple kinetic characteristics. However, this approach also limits the analysis of electrochemical interface performance in this study. Further research on this topic should incorporate EIS and porosity as investigative tools for electrochemical interface characteristics.