Introduction

Nowadays, exploration of a new anode material with high safety and excellent cyclability, as compared to commercial carbon/graphite materials, has become the key issue in the development of the electric vehicle industry [1].

Spinel Li4Ti5O12 is regarded as an alternative anode material for lithium-ion batteries due to its appealing features such as “zero-strain” structure characteristic, stable operating voltage (1.55 V vs. Li/Li+), good cycling stability, safety, and simple synthesis [2–4]. Therefore, a single cell with an operating voltage of about 2.5 V can be provided when Li4Ti5O12 couples with high-voltage cathode materials such as LiCoO2, LiMn2O4, Li[Li0.1Al0.1Mn1.8]O4, and LiNi1/3Mn1/3Co1/3O2, meaning 12 V batteries can be made connecting five cells in series [5–8] so that the batteries can be used as power sources for hybrid electric vehicles (HEV) or plug-in HEV.

Despite many advantages associated with Li4Ti5O12, low electric conductivity prohibits its use in large-scale applications. Therefore, many strategies have been employed to enhance the electric conductivity of Li4Ti5O12. Compared with other optimization methods, carbon doping has been proved as an effective and feasible way. Luo et al. [9] synthesized Li4Ti5O12/C nanorods by a wet-chemical route using glucose as carbon source. Chen et al. [10] prepared Li4Ti5O12-KB composites by ultrasonication-assisted sol-gel method using Ti(OC4H9)4, LiCH3COO · 2H2O, and conductive carbon black as the raw materials. Ni et al. [11] prepared nano-Li4Ti5O12/CNT composites by a liquid-phase deposition method using acetic acid as the chelating agent. In addition, Li4Ti5O12/C derived from other carbon sources such as organic acids [12, 13] and polymers [14, 15] were examined. However, previous studies [9–15] mainly focused on the synthesis of Li4Ti5O12/C composite materials by adding extra carbon sources. The extra carbon sources are added, resulting in relatively complex synthesis process and inconsistency in product quality.

In this work, spinel Li4Ti5O12/C composites were synthesized by one-step solid-state reaction using lithium lactate (C3H5O3Li) as lithium and carbon sources and TiO2 as a titanium source. The facile one-step synthesis can reduce production steps, contribute to uniform mixing of raw materials, and get homogenous Li4Ti5O12/C composites. Moreover, as-prepared Li4Ti5O12/C composites with low in situ carbon content of only 1.81 wt% show significant improvement in discharge capacity and high rate cycling performance comparable to Li4Ti5O12/C prepared by solid-state reaction in previous studies.

Experimental

The Li4Ti5O12/C composites were synthesized from a stoichiometric amount of TiO2 and C3H5O3Li as raw materials. The raw materials were mixed in ethanol as disperser and then ball-milled for 6 h. Subsequently, the mixtures were dried and sintered at 300, 600, and 800 °C for 1, 2, and 8 h in flowing nitrogen, followed by cooling down to room temperature. The final Li4Ti5O12/C composites were obtained by grinding. As a comparison, pristine Li4Ti5O12 was prepared using similar method but heated in air.

Electrochemical properties of the Li4Ti5O12/C composites were evaluated using 2032 coin cells with a lithium metal foil as the counter electrode. The Li4Ti5O12/C working electrodes were made by mixing active materials, conductive carbon black, and an aqueous binder LA132 (Indigo, China) in a weight ratio of 85:10:5 and then pasted uniformly onto a copper foil and dried at 80 °C under vacuum for 4 h to give the electrodes. The cells were assembled in an argon-filled dry glove box. The electrolyte was a 1.0 mol L−1 LiPF6 solution in the mixture of ethylene carbonate, dimethyl carbonate, and ethylene methyl carbonate (1:1:1 by volume). A Celgard 2400 polypropylene membrane was used as the separator.

Thermogravimetric (TG) analysis was performed by a Shimadzu TA-50 thermal analyzer with a heating rate of 10 °C min−1 from 30 to 950 °C at a nitrogen flow rate of 50 mL min−1. X-ray diffraction (XRD) was carried on a Shimadzu XRD-7000S X-ray diffractometer with Cu-Kα radiation of λ = 1.5418 Å in the range of 10° < 2θ < 90°. Scanning electron microscopy (SEM) was characterized on a FEI Quanta 450 scanning electron microscope. Transmission electron microscopy (TEM) was revealed by a FEI Tecnai G20 transmission electron microscope. The amount of carbon in the Li4Ti5O12/C composites was measured by a EuroVector EA3000 elemental analyzer. Raman spectra were recorded using a confocal Micro-Raman spectroscopy fitted with a 632.8-nm He-Ne excitation laser. Electrochemical impedance spectra (EIS) were performed by an Autolab PGSTAT 128N in the frequency range from 10−1 to 106 Hz. Constant current charge-discharge and cycle performance of the cells were tested on a land battery tester and cyclic voltammograms were investigated by an Autolab PGSTAT 128N. All the tests were carried out at room temperature.

Results and discussion

The TG-DTG curves of the precursor mixtures consisting of C3H5O3Li and TiO2 are shown in Fig. 1. A small peak at 141.69 °C is observed, relating to the removal of adsorbed water. A continuous weight loss between 300 and 600 °C can be related to the thermal decomposition of the precursor. There are two obvious peaks at 301.69 and 456.69 °C, which is relevant to the thermal decomposition of C3H5O3Li. When the temperature exceeds 700 °C, the TG curve shows nearly constant weight, indicating that the reaction is complete and well-crystallized Li4Ti5O12 is obtained. Currently, Li4Ti5O12/C composites (or Li4Ti5O12) were synthesized by high-temperature solid-state reaction using TiO2 as titanium source and different lithium sources. Different lithium sources lead to small differences in decomposition temperatures of solid-state reaction (DTSSR). Inorganic lithium sources include Li2CO3 (DTSSR >460 °C) [16, 17], LiOH (DTSSR >500 °C) [18], and LiNO3 (DTSSR >560 °C) [19]. Organic lithium compounds include LiAc (>470 °C) [20] and LiOAc (>320 °C) [21]. Polymer lithium sources include PAALi (>420 °C) [22]. Compared with the above lithium sources, C3H5O3Li shows lower DTSSR, indicating that the synthesis route using C3H5O3Li as lithium source consumes less energy which is beneficial to industrial production and worthy of further investigation.

Fig. 1
figure 1

TG-DTG curves of the Li4Ti5O12/C precursor

Full size image

Figure 2 presents the XRD patterns of the Li4Ti5O12/C composites and pristine Li4Ti5O12. It can be seen that all the diffraction peaks between 10 and 90 °C can be indexed to the standard Li4Ti5O12 pattern (PDF no. 49–0207), demonstrating the obtained samples have a cubic spinel phase structure with no evidence of impurities. This means that the carbon derived from C3H5O3Li does not affect the structure of Li4Ti5O12 during the heat treatment. The unchanged structure characteristic of Li4Ti5O12/C ensures its excellent electrochemical cycling performance as pristine Li4Ti5O12. In addition, there are no obvious diffraction peaks of the carbon in XRD patterns because of its low content or amorphous state.

Fig. 2
figure 2

XRD patterns of the Li4Ti5O12/C composites and pristine Li4Ti5O12

Full size image

Figure 3a, b shows the SEM images of the as-prepared Li4Ti5O12/C and Li4Ti5O12 particles. As seen, two samples display regular crystalline morphology. The particle size of Li4Ti5O12/C is much smaller than that of Li4Ti5O12, which exhibits that the in situ carbon from the pyrolysis of C3H5O3Li can effectively prevent the growth of particles during sintering process. And the smaller particles of Li4Ti5O12/C are more likely to fuse together into larger agglomerates. Usually, it is considered that a smaller Li4Ti5O12 particle size will shorten lithium-ion diffusion path and contribute to better electrochemical properties. Moreover, the carbon materials can improve electrode performances from some aspects. However, on the other hand, the higher the inactive materials (carbon) content is, the lower the discharge capacity is. It is very important to decrease the inactive materials to balance electrode capacity and other performance for practical use [23, 24]. Currently, the common in situ carbon content in the prepared Li4Ti5O12/C composites via solid-state reaction is no less than 3 wt% [25–28]. In this work, the content of in situ carbon in the Li4Ti5O12/C composites is 1.81 wt% measured by elemental analysis. Because the in situ carbon content is lower than those reported in previous studies, it is expected that the discharge capacity and related performances of the Li4Ti5O12/C composites are better than those in previous studies. In order to observe the detail distribution of the carbon and further verify the particle size and microstructure, the composites were closely studied by TEM. As shown in Fig. 3c, it can be clearly seen from the micrograph that the Li4Ti5O12 particles are wrapped with an amorphous carbon web, which is consistent with the conclusions of XRD and elemental analysis. Combined with the before-mentioned morphology analysis, the Li4Ti5O12/C composites are expected to give excellent electrochemical performances.

Fig. 3
figure 3

SEM image of the Li4Ti5O12/C (a) and Li4Ti5O12 (b) and TEM image (c) of the Li4Ti5O12/C

Full size image

Figure 4 shows the Raman spectra of the Li4Ti5O12/C composites. The intense broadbands at about 1325 and 1591 cm−1 are assigned to the disorder (D) and graphene (G) bands of in situ carbon in Li4Ti5O12/C composites, respectively. The G band corresponds to one of the E2g modes, which has been assigned as the sp2 graphite-like structure, whereas the D band corresponds to one of the A1g modes, which is attributed to the sp3 type tetrahedral carbon. The value I D/I G (the peak intensity ratio) can be used to evaluate the content of sp3− and sp2− coordinated carbon in the sample, as well as the degree of disorder for the in situ carbon. Theoretically, the high values for the I D/I G and A sp3/A sp2 (the area ratio of sp3− to sp2− coordinated carbon) parameters indicate a low degree of graphitization [9]. The value of the I D/I G ratio of 1.17 is calculated. It suggests that the in situ carbon is mainly in an amorphous structure, which is consistent with the conclusions of TEM.

Fig. 4
figure 4

Raman spectra of the Li4Ti5O12/C composites

Full size image

Figure 5 shows the cyclic voltammograms (CVs) of the as-prepared Li4Ti5O12/C and Li4Ti5O12 electrodes. It is obvious that both samples show a pair of sharp redox peaks, which are the characteristics redox peaks of Li4Ti5O12. At a scan rate of 0.2 mV s−1, it should be noted that the redox peak profile is located at 1.46/1.68 V for the Li4Ti5O12/C, and 1.36/1.73 V for the Li4Ti5O12. The Li4Ti5O12/C electrode has a smaller potential separation compared with the Li4Ti5O12 electrode, showing a smaller polarization. This could be attributed to the following reasons. First, the smaller size makes shorter distance for lithium-ion diffusion in Li4Ti5O12/C particles and benefits the diffusion process. Moreover, the in situ carbon in Li4Ti5O12/C composites effectively improves the electrical contact of each Li4Ti5O12 particle, thus enhancing electrochemical performances of Li4Ti5O12. These results are consistent with the conclusions of SEM and TEM micrographs.

Fig. 5
figure 5

CVs of the Li4Ti5O12/C and Li4Ti5O12 electrodes

Full size image

Figure 6 presents EIS of Li4Ti5O12/C and Li4Ti5O12 cells. It can be seen that the Li4Ti5O12/C cell exhibits much lower impedance than the Li4Ti5O12 cell, indicating much better electronic conductivity of the Li4Ti5O12/C composites. According to the formula σ = d/AR (where σ is the conductivity, R is the resistance, d is the thickness, and A is the specific surface area) [29], the electronic conductivities of Li4Ti5O12 and Li4Ti5O12/C are 8.4 × 10−8 and 1.8 × 10−7 S cm−1, respectively. The increase of the electronic conductivity in the Li4Ti5O12/C composite is mainly due to the presence of pyrolytic carbon. As a result, the redox polarization decreases from 40 to 37 mV based on the charge-discharge curves at 0.2 C. Cycled at 20 C, the insertion/extraction polarization reduces from 532 to 373 mV as shown in Fig. 7. These facts are consistent with the results of CVs. Furthermore, with the increase of current density, the electrode polarization increases gradually and the discharge capacity gradually decreases as shown in Fig. 7. At a low rate of 0.2 C, the initial discharge capacity of 168.4 mAh g−1 is delivered for the sample (it is 171.5 mAh g−1 if the capacity is calculated based on the weight of the Li4Ti5O12 in the Li4Ti5O12/C electrode), which is close to its theoretical capacity of 175.0 mAh g−1. At high current rates of 10 C and 20 C, the samples still show 140.0 and 133.3 mAh g−1, respectively. Some results of previous works on the Li4Ti5O12/C prepared by solid-state reaction are listed in Table 1. When total carbon content is similar, the discharge capacity of Li4Ti5O12/C using C3H5O3Li, especially at high rates, is better than the results listed in Table 1. It was worth noting that carbon-free Li4Ti5O12 electrode showed excellent electrochemical performance both in the rate capability and the long-term cyclability, and the electrode delivers 9 % higher capacity than carbon composite electrode in ref. [31]. The idea of ref. [31] is essentially consistent with our concept that very low carbon content in the Li4Ti5O12 electrode. Does Li4Ti5O12 need carbon in lithium-ion batteries? The answer is yes. Firstly, the high rate capability of Li4Ti5O12 is one of its most important advantages as an anode material. As shown in Fig. 3a (ref. [31]), the discharge capacities of the carbon-containing Li4Ti5O12 electrode gradually exceeded those of the carbon-free Li4Ti5O12 electrode when charge-discharge rate reached 3 C. With the increase of the charge-discharge rate, reversible capacity decay of the carbon-containing Li4Ti5O12 electrode was significantly less than that of the carbon-free Li4Ti5O12 electrode, which illustrated that the carbon helped to improve high rate discharge capacity. Secondly, high coulombic efficiency is important for the electrode materials. As shown in Fig. 3a (ref. [31]), with the increase of the charge-discharge rate, coulombic efficiencies of the carbon-containing Li4Ti5O12 electrode were significantly higher than those of the carbon-free Li4Ti5O12 electrode, which indicated that the carbon helped to improve high rate cycling performance. Compared with ref. [31], further research has been done in our work. At high rates (>2 C), the discharge capacities of Li4Ti5O12/C electrode are significantly higher than those of ref. [31], which is due to the presence of carbon in the as-prepared Li4Ti5O12/C. Moreover, Li4Ti5O12/C electrode at high rates provides high coulombic efficiencies close to 100 %, which is significantly higher than those of ref. [31] (see Fig. S4 in the reference. 2 C: 94.0 %a, 5 C: 83.0 %a, 10 C: 72.0 %a). These results show that the Li4Ti5O12/C composites have high rate capacity, which is due to the low in situ carbon content in them, indicating that lithium lactate is indeed suitable lithium and carbon sources for the one-step solid-state synthesis of Li4Ti5O12/C composites. Moreover, it can be expected that the Li4Ti5O12/C composites provide better high rate capability because of its higher electronic conductivity compared with the pristine Li4Ti5O12.

Fig. 6
figure 6

EIS of the Li4Ti5O12/C and Li4Ti5O12 cells

Full size image
Fig. 7
figure 7

The first charge-discharge curves of the Li4Ti5O12/C composites (a) and pristine Li4Ti5O12 (b) at different rates

Full size image
Table 1 Discharge capacities, carbon contents, and capacity retention of Li4Ti5O12/C prepared by solid-state reaction in the reference works
Full size table

The rate performances of the Li4Ti5O12/C composites and pristine Li4Ti5O12 at different rates are shown in Fig. 8. It can be seen that the Li4Ti5O12/C composites have high rate performance though the reversible capacities decreased gradually with the increase of current density. Moreover, the Li4Ti5O12/C composites demonstrate obviously improved reversible capacity and rate performance over the pristine Li4Ti5O12, especially at higher current rates. Even at 20 C, the discharge capacity still reaches 109.8 mAh g−1 after 50 cycles, while for pristine Li4Ti5O12, the corresponding value is 43.5 mAh g−1. The results show that the Li4Ti5O12/C composites exhibit excellent rate performance, which is attributed to the amorphous carbon layer on the surface of Li4Ti5O12. The pyrolytic carbon webs derived from C3H5O3Li increase the electronic conductivity of Li4Ti5O12, and hence improve its electrochemical properties, especially the high rate capability.

Fig. 8
figure 8

The rate performance of the Li4Ti5O12/C composites (a) and pristine Li4Ti5O12 (b) at different rates

Full size image

Figure 9 shows the cycling performance of the Li4Ti5O12/C composites and pristine Li4Ti5O12 at 1 C and 10 C. The Li4Ti5O12/C composites and pristine Li4Ti5O12 both provide high coulombic efficiencies close to 100 %. At a low rate of 1 C, the initial capacity is 168.0 mAh g−1 with capacity retention of 94.4 % at 200th cycle for the Li4Ti5O12/C composites, while for pristine Li4Ti5O12, the corresponding values are 136.6 mAh g−1 and 87.2 %, respectively. At a high rate of 10 C, the initial capacity is 140.0 mAh g−1 with capacity retention of 90.2 % at 200th cycle for the Li4Ti5O12/C composites, while for pristine Li4Ti5O12, the corresponding values are 114.3 mAh g−1 and 62.5 %, respectively. Compared with previous studies listed in Table 1, the Li4Ti5O12/C composites in this work exhibit excellent low rate cycling performance the same as those in previous studies. Furthermore, their high rate cycling performance is better than those of Li4Ti5O12/C in previous studies. The results show that the as-prepared Li4Ti5O12/C composites exhibit excellent high rate cycling performance, which is attributed to the presence of pyrolytic carbon. The abovementioned electrochemical facts illustrate that low in situ carbon content of the as-prepared Li4Ti5O12/C composites perfectly balances discharge capacity and high rate cycling performance.

Fig. 9
figure 9

The cycle curves of the Li4Ti5O12/C composites and pristine Li4Ti5O12 at 1 C and 10 C

Full size image

Conclusions

Spinel Li4Ti5O12/C composites were obtained from C3H5O3Li and TiO2 via one-step solid-state reaction without adding extra carbon sources. Herein, the commonly used organic lithium salt C3H5O3Li supplies both lithium and carbon sources. The low in situ carbon content of the as-prepared Li4Ti5O12/C composites perfectly balances discharge capacity and high rate cycling performance. At 0.2 C, the initial discharge capacity of 168.4 mAh g−1 is delivered for the sample, which is close to its theoretical capacity of 175.0 mAh g−1. At 1 C and 10 C, the initial capacity is 168.0 and 140.0 mAh g−1 with capacity loss of 5.6 and 9.8 % at 200th cycle, respectively. Even at 20 C, the discharge capacity still reaches 109.8 mAh g−1 after 50 cycles. As a result, the Li4Ti5O12/C composites derived from this synthetic route show high discharge capacity and excellent rate cycling performance, which makes them a promising anode candidate for electric vehicles.