Abstract

The spinel Li4Ti5O12 (LTO) has been doped by Ca2+ via a solid-state reaction route, generating highly crystalline Li3.9Ca0.1Ti5O12 powders in order to improve the electrochemical performance as an anode. The structure changes, morphologies, and electrochemical properties of the resultant powders have been characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), and cyclic voltammetry (CV), respectively. Crystal structure and composition were analyzed, and results were obtained with various tests of LTO. Electrochemical measurements revealed that Li3.9Ca0.1Ti5O12 anodes exhibit better rate capability, better cycling stability, and a higher specific capacity than pure LTO anodes.

1. Introduction

Lithium-ion batteries (LIBs) have been widely utilized as the most important next-generation energy sources [1, 2]. However, the carbon anode materials are facing safety issues caused by relatively low Li-intercalation potential (about 0 V vs. Li/Li+). As an alternative anode material for LIBs, spinel Li4Ti5O12 (LTO) has attracted more and more attention because of its extremely small volume change during Li+ insertion/extraction as the so-called “zero strain material” [3]. At the same time, LTO possesses the fast Li+ insertion and deinsertion ability, excellent cycle stability, and high thermal stability due to its voltage plateau at 1.55 V (vs. Li/Li+), which can inhibit the decomposition of the electrolyte on the electrode/electrolyte surface and hence reinforce the safety of LIBs [4]. Unfortunately, LTO exhibits a poor electronic conductivity [5]. To improve the conductivity, a considerable number of effective ways have been developed. Generally, the strategy includes the preparation of nanometer particles [6], carbon coating [7], and doping with aliovalent metal ions [815] (V5+ [8], Sr2+ [9], La3+ [10], Mg2+ [11], Al3+ [12], Co3+ [13], and Ru3+ [14]) in Li, Ti, or O sites; the selection of morphologies has a crucial influence on the electrochemical characteristics of LTO materials [1518]. Additionally, there are literatures which reported various strategies including combined ion coping or sheet-encapsulated coating [12, 1922]. In this work, the effects of Ca-doping on the microstructure, morphology, and electrochemical characteristics of LTO have been investigated. The doped Ca2+ can replace the Li+ of 8a to produce free electrons, thus enhancing the conductivity of the material. The results indicate that the Ca-doped LTO (Li3.9Ca0.1Ti5O12) delivers a large capacity at a high rate with excellent cycle performance, which provides a potential technique for the application of LTO in LIBs with high performance.

2. Materials and Methods

2.1. Preparation of the Samples

Ca-doped LTO (Li3.9Ca0.1Ti5O12) and undoped Li4Ti5O12 were generated via solid-state reaction route. CaO was used as a calcium source for doping. Excess Li (2 wt%) was necessary because of the loss of Li during synthesis at high temperature. A mixture of TiO2, Li2CO3, and CaO in a proper amount in high-energy ball milling for half an hour; then, the powders were put in a continuous heating tube furnace at 800°C for 12 h in a flowing Ar atmosphere.

2.2. Preparation of the Batteries

In order to prepare the electrode, LTO or Ca-doped LTO, super carbon, and polyvinylidene fluoride (PVDF) (LTO/C/PVDF in a weight ratio of 8 : 1 : 1) were uniformly mixed in N-methyl pyrrolidone (NMP) solvent to produce homogeneous slurry, which was cast on a copper foil by a blade. The solvent was evaporated under vacuum at 110°C overnight. The electrode was punched to a disk shape with a diameter of 16 mm. Half cells were assembled in a glove box with metallic lithium foils as the counter electrode, microporous polyethylene membranes (Celgard 2400) as the separators, and 1 M LiPF6 in a mixture of ethyl carbonate (EC) and dimethyl carbonate (DMC) (EC/DMC in a volume ratio of 1 : 1) as the electrolyte [1922].

2.3. Material Characterization

The structure and morphology of the Ca-doped LTO powders were characterized by X-ray diffraction measurement (XRD, PANalytical, X’Pert Powder) and scanning electron microscopy (SEM, Hitachi, SU8010). Galvanostatic cycling was conducted on a Wuhan LAND Electronics Battery Tester. Cyclic voltammetry (CV) was tested on an electrochemical workstation (Autolab, PASTAT101) between 0.8 and 2.8 V (vs. Li+/Li) at a scanning rate of 0.5 mV s−1. Electrochemical impedance spectrum (EIS) measurements were also tested using Autolab at a potentiostatic signal amplitude of 5 mV and with a frequency range from 10 MHz to 1 MHz [23].

3. Results and Discussions

XRD patterns of the obtained LTO and Ca-doped LTO are shown in Figure 1(a). The diffraction peaks of all the samples can be completely indexed in the cubic spinel structure (JCPDS#49-0207). No impurity peaks were observed, which indicates that the Ca has entered the lattice structure of LTO without changing its crystal structures. And also, it is hypothesized that the narrow width of the Ca-doped LTO peaks is caused by the high crystallinity of the powders [24].

For clear observation, enlarged (111) peaks of LTO and Li3.9Ca0.1Ti5O12 are shown in Figure 1(b). It can be observed that the (111) peak shifted to low angle with Ca-doping, suggesting that the doped LTO samples possess larger lattice constant than pristine LTO. This result could be attributed to Ca-doping which makes the phase structure change from a cubic structure to an oblique crystal structure and opens up a channel favorable for lithium-ion migration along the axis. The substitution of Li+ with the valence of Li+ is the formation of Ti4+/Ti3+-mixed valence state at B site based on charge compensation. As a result, the lattice constant was only slightly enlarged. Theoretically, the lattice stability could be maintained during the fast lithium-ion transfer, which would improve the performance of Ca-doped LTO electrodes in cell testing.

SEM images of LTO powders with and without Ca-doping are illustrated, respectively, in Figure 2. Both of the two materials possess uniform particle size distribution. Generally, the crystallite size is ranged from 100 to 200 nm in both samples. This fine microstructure could be attributed to the well dispersion and fully reaction of TiO2, Li2CO3, and CaO. It can be explained that the small particle size is beneficial for decreasing the interface impedance between the active materials and electrolytes which thus promotes the efficiency of Li+ insertion/extraction into the LTO crystal structure [25].

In order to investigate the effect of Ca-doping on improving electrochemical properties, cyclic voltammetry potentiostatic (CV) of Li3.9Ca0.1Ti5O12 and LTO electrodes have been measured at a scan rate of 0.5 mV s−1 between 0.8 V and 2.8 V (Figure 3). It is obvious that there is a pair of reversible redox peaks for each sample, which represents the lithium insertion process (reduction peaks at about 1.50 V vs. Li+/Li) and extraction process (oxidation peaks at about 1.75 V vs. Li+/Li). It is attributed to the typical two-phase reaction of the redox couple of Ti3+/Ti4+ during Li+ insertion/extraction processes. The corresponding electrochemical reaction is Li4Ti5O12 + 3Li+ + 3e − →Li7Ti5O12 ( = 1.55 V) [26].

The rate capability of the Ca/LTO and LTO electrodes at different rates is presented in Figure 4. It can be observed that the Ca-doped LTO exhibited a higher discharge capacity and good cycling stability at a low rate, and its initial discharge capacity could reach up to 155.7 mAh g−1 when at 0.1C. As the discharge rate increases, the capacity of undoped LTO quickly decreases. However, the Ca-doped LTO exhibits less capacity degradation. At 5C, the discharge specific capacity of LTO and Ca/LTO was 68.7 mAh g−1 and 106.1 mAh g−1, respectively. Compared with the first discharge capacity, the retention rate was 44.12% and 68.01%, respectively. The better rate capability and cycling stability of Ca-doped LTO were observed because of the efficient Li+ insertion/extraction induced by Ca-doping.

The cycling performance of the LTO and Ca/LTO samples at rates of 2C and 5C is shown in Figure 5. As expected, Ca/LTO electrodes exhibited higher reversible capacities than that of undoped LTO. At 2C, the discharge capacity of Ca/LTO remained 115.1 mAh g−1 even after 100 cycles (Figure 5(a)). Compared with the first discharge specific capacity of 121.7 mAh g−1, the attenuation amplitude was only 5.42%. In contrast, the corresponding value of the LTO was decreased to 80.1 mAh g−1. The attenuation was 13.87%. When at 5C rate (Figure 5(b)), the initial discharge capacity of Ca/LTO was 107.3 mAh g−1 and remained as high as 97.7 mAh g−1 after 100 cycles (91.05% retention), while the initial discharge capacity of LTO was 75.2 mAh g−1 and remained 61.4 mAh g−1 after 100 cycles. These results demonstrate that the Li+ insertion/extraction processes of Ca/LTO exhibits better reversibility and cyclic stability even at high rates and higher capacity retention than pure LTO, which should be attributed to the relatively slight change of lattice volume in Ca-doped LTO electrodes during the insertion and extraction of Li+.

To evaluate the electrochemical performance of Ca-doped LTO electrodes in LIBs, galvanostatic charge-discharge tests have been carried out. Figure 6 shows the initial charge-discharge curves of the LTO and Ca/LTO samples at various current rates from 0.1C to 5C [27]. It is obvious that the discharge plateau of each material decreases with increasing current rate, because of the increased polarization of electrodes. The Ca/LTO delivered relatively higher discharge-specific capacities especially at high current rates.

Figure 7 shows the EIS curves of cells with LTO and Ca/LTO at the voltage of 1.55 V after 20 cycles. The Ca/LTO electrode exhibits relatively lower resistance than pure LTO electrode, which is consistent with the result of the better reversible capacities. In the EIS patterns, a semicircle in the high-frequency region and a straight line in the low-frequency region are observed. The semicircle is related to the charge transfer resistance between the active material interface and the electrolyte, while the straight line represents the Warburg impedance that is related to diffusion of Li+ [2830]. In the equivalent circuit, is the Warburg impedance, and and represent the ohmic resistance and the charge transfer resistance, respectively [2032]. The diffusion coefficient values of the lithium ions in the bulk electrode materials were calculated from EIS data demonstrating that the charge transfer process of Li+ on the surface of Ca/LTO electrodes has been significantly improved and ion conductivity of Ca/LTO electrodes has been promoted via Ca-doping.

4. Conclusions

Generally, the Ca-doped LTO can be synthesized by solid-state methods and the electrochemical performance of Ca-doped LTO is much better than the pure LTO. There was no significant change in the particle morphology of LTO powders after Ca-doping, and the small particle size is beneficial for decreasing the interface impedance between the active materials and electrolytes which thus promotes the efficiency of Li+ insertion/extraction into the LTO crystal structure. The rate capability, cyclic stability, and the ion conductivity of the half cells with Ca-doped LTO anodes have been significantly improved, compared to cells assembled with the pure LTO, which should be attributed to the relatively slight change of lattice volume in Ca-doped LTO electrodes during the insertion and extraction of Li+. The excellent performance of Ca-doped LTO anodes can be attributed to its promoted charge transfer of Li+ and relatively lower interfacial resistance arising from Ca-doping. The Ca/LTO delivers relatively higher discharge-specific capacities especially at high current rates. The charge transfer process of Li+ on the surface of Ca/LTO electrodes can be significantly improved, and ion conductivity of Ca/LTO electrodes has been promoted via Ca-doping. All the results demonstrate that the Ca-doped LTO materials could be potential in the application of safe LTO anodes in high rate LIBs.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare no conflict of interest.

Acknowledgments

This project was supported by the fund from China NSF (51703081), Wuhan Applied Basic Research Project (2018010401011285), and 4th Yellow Crane Talent Programme of Wuhan City (08010004).