Introduction

Among the currently available energy storage technologies, lithium-ion (Li-ion) battery has the highest energy density and, hence, has received intense attention from both the academic community and industry as the power source in hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), and full electric vehicles (EVs). However, the large-scale deployment of Li-ion batteries to power the automotives is significantly hindered by a few major technological barriers, including high cost, insufficient cycle life, intrinsically poor safety characteristics, and poor low-temperature performance (<−20 °C). Li-ion batteries with both high power and energy densities are required to meet the ever increasing demands in these key areas [1]. A conventional Li-ion cell usually has the anode based on carbon, the cathode a metal oxide such as LiCoO2, LiNiO2, and LiMn2O4 and a Li+ ion salt in organic solvents as the electrolyte. During the charging process, Li+ ions are extracted from the metal oxide cathode and get inserted into the carbon anode. During the discharging process, the electrode reaction is reversed. The commercial Li-ion batteries can deliver a voltage of 3.6 V with a low self-discharge. Recent research works on Li-ion cells mainly focus on the methods to improve suitably the safety features and electrochemical performance of electrode materials [2].

Carbon in the form of graphite is the most widely used anode material in Li-ion cells. The electrochemical activity of graphite is based on the intercalation and deintercalation of Li+ ions into and from its layered structure [3]. Although the Li-ion cell with graphite as anode and a suitable cathode can lead to high energy density, its rate capability is limited due to the slow solid-state diffusion of Li+ ion within the electrode materials. Li+ insertion into graphite occurs at a potential of <1 V vs. Li+/Li (Fig. 1). At such low potentials, reduction of the organic electrolyte occurs, leading to the formation of a passivating solid electrolyte interface (SEI) layer on the surface of graphite during the first few charge–discharge cycles [4]. The formation of SEI is essential for the operation of graphite electrodes, as it inhibits further reaction of electrolyte at the electrode surface. However, the SEI layer impedes the insertion/removal of Li+ ions during subsequent cycles and leads to poor cycle performance of the battery. Also, formation of SEI leads to irreversible capacity loss during initial cycles [5].

Fig. 1
figure 1

First cycle charge–discharge curve of graphite vs. Li (reproduced with permission from [5])

Full size image

Graphite undergoes volumetric expansion and contraction during the cell charging and discharging, respectively, which results in peeling off of active material from the current collector at high rate and leads to short circuit or capacity fading. The performance gets further affected under subzero as well as high operating temperatures. The graphite-based electrodes lack inherent safety because Li+ intercalation occurs at a potential close to that of lithium, and there is no end-of-charge indicator in the voltage profile that can signal the onset of catastrophic oxygen evolution from the cathode materials due to overcharge [6–8].

To overcome the disadvantages associated with the graphite/carbon-based anode, the use of an electrochemical redox couple that intercalates/deintercalates Li+ at a higher potential was proposed. Lithium titanate (Li4Ti5O12 abbreviated as LTO) with the Ti4+/Ti3+ redox couple working at approximately 1.55 V vs. Li fulfills this requirement [9–12]. This metal oxide has been extensively studied during the last few years and found very promising as anode material in Li-ion cells.

Characteristics of LTO

LTO is a “zero strain” material in terms of Li+ intercalation and deintercalation with only a slight shrinkage of the lattice parameter, from 8.3595 to 8.3538 Å resulting in a minimal change in cell volume (about 0.2 %). Since the Li+ ions have the same size as the sites they occupy in the crystal structure of LTO, the LTO particles do not have to expand or shrink substantially when the ions are entering or leaving the structure (change in ionic radius of Ti by redox reaction might contribute to slight volume change). This property helps to maintain the structural stability of Li4 + x Ti5O12 with minimal particle fatigue during charge–discharge processes and results in a cell that can provide extremely long cycle life and good rate capability compared to conventional Li-ion cells [13–17]. The zero strain nature of LTO was experimentally evaluated by Choi et al. [18]. They carried out the stress evolution studies on LTO for various voltages and lithiation/delithiation rates. As per their studies, the features in the stress signal are correlated to the coexistence of the two phases Li4Ti5O12 and Li7Ti5O12 as well as to structures that are formed during under- and overlithiation. They also showed that large volume expansion occurs particularly below 1 V vs. Li/Li+. Upon lithium insertion, the lithium ions move from the 8a to the 16c sites, and at a composition of Li7Ti5O12, a rock salt structure is formed whereby all 16c sites are occupied and all 8a sites are empty. This defective structure has an increased number of vacancies at the 8a site, which facilitate lithium diffusion through 16c → 8a → 16c pathways [19]. Most of the studies showed that LTO forms a relatively clean interface in LiPF6-based carbonate electrolytes. It does not form a SEI layer akin to that formed on graphite electrodes because its 1.55 V plateau vs. Li is higher than the reduction potential of carbonate solvents [20]. Figure 2 depicts the typical charge–discharge curve of LTO vs. Li [21]. LTO is normally cycled in the voltage range 1.0–2.5 V vs. Li/Li+.

Fig. 2
figure 2

Typical charge–discharge curve of LTO vs. Li (reproduced with permission from [21])

Full size image

One mole of LTO can accommodate 3 mol of Li+ as per the equation given below, resulting in a theoretical specific capacity of 175 mAh g−1 [20]. The capacity is lower than that of graphite (372 mAh g−1) which takes up one Li+ per six carbon atoms. This is due to the lower electrochemical equivalence of LTO (molecular weight, 457 g mol−1) compared to graphite (molecular weight, 72 g mol−1).

$$ {\mathrm{Li}}_4{\mathrm{Ti}}_5{\mathrm{O}}_{12+}3{\mathrm{Li}}^{+}+3{\mathrm{e}}^{-}\underset{\mathrm{charge}}{\overset{\mathrm{discharge}}{\rightleftarrows }}{\mathrm{Li}}_7{\mathrm{Ti}}_5{\mathrm{O}}_{12} $$

LTO has a face-centered cubic spinel structure [22–24] as shown in Fig. 3. The 32e positions are occupied by oxygen atoms, 5/6 of the 16d positions are occupied by Ti atoms and the rest of the 16d positions by Li atoms, forming a stable [Li1/3Ti5/3]16dO4 framework with a space group of Fd3̅m. The tetrahedral (8a) sites are occupied by Li atoms while the octahedral (16c) sites are empty, and the structure is represented as [Li]8a[Li1/3Ti5/3]16d[O4]32e. LTO has empty Ti 3d states with band gap energy of 2–3 eV, which gives an insulating character to this material.

Fig. 3
figure 3

Crystal structure of LTO (reproduced with permission from [23])

Full size image

The typical XRD pattern and Raman spectrum of LTO are shown in Figs. 4 and 5, respectively. Sharp diffraction peaks are observed at 2θ = 18.5°, 35.7°, 43.4°, 47.5°, 57.3°, 62.9°, 66.2°, 74.5°, 75.5°, 79.5°, and 82.4° corresponding to (111), (311), (400), (331), (333), (440), (531), (533), (622), (444), and (551) planes, respectively, of spinel structure [21]. In the Raman spectrum, the two strong bands at 233.9 and 672.4 cm−1 are assigned to the characteristic modes of A1g and F2g. The medium strength band at 423.1 cm−1 corresponds to the Eg mode, and the weaker bands at 270.5 and 341.5 cm−1 belong to F2g [21, 25–32].

Fig. 4
figure 4

XRD pattern of LTO (reproduced with permission from [21])

Full size image
Fig. 5
figure 5

Raman spectra of LTO (reproduced with permission from [21])

Full size image

Synthesis methods

The electrochemical performance of the electrode materials used in Li-ion cells is strongly affected by the properties, such as purity and composition of the material, morphology, specific surface area, and crystallinity, which in turn are highly dependent on the synthesis methods adopted. LTO has been synthesized through a variety of techniques, including solid state, sol–gel, hydrothermal, microwave, combustion, sonochemical, molten salt, and spray pyrolysis.

For solid-state synthesis of LTO, researchers have employed varying sintering conditions, mostly 800–1,000 °C for 10–24 h [33–39]. Several studies have shown that fine particle-sized LTO exhibits excellent rate capability due to shorter length for Li+ diffusion and electron transfer [40–46]. The heat treatment atmosphere also affects the rate capability of LTO. According to Yuan et al. the optimal calcination atmospheres for the synthesis of LTO and LTO/C were diluted hydrogen (H2/Ar) and nitrogen, respectively. The oxygen vacancy created due to the reduction of TiO2 in the reducing atmosphere of hydrogen helps to enhance the oxygen ion diffusivity during calcinations, whereas carbon coated over TiO2 in the starting material acts as reducing agent during calcination in nitrogen atmosphere. Thus, the LTO samples prepared by heating under H2/Ar or N2 atmosphere showed a remarkably higher rate capability and better cycle stability compared to samples prepared by heating in air [47–50]. The main disadvantage of solid-state synthesis is that it generally does not provide the required small particle size, morphology, and homogeneity, which are desirable for better electrochemical performance of the final product [23]. Ball milling-assisted solid-state method is an effective route for the synthesis of well-defined mesoporous, ultrafine nanoparticles, which can deliver high capacity even at 10 C [51]. The mesoporous LTO obtained through this method delivered a high initial discharge capacity of 174.5 mAh g−1 at 10 C rate, which gets retained at 143.4 mAh g−1 even after 50 cycles. The excellent high-rate discharge and cycle performance for the LTO prepared by the ball milling-assisted solid-state method can be attributed to its ultrafine particle size and the presence of mesoporous structure. The mesopores can facilitate the fast transport of solvated electrolyte and lithium ions throughout the electrode materials, whereas the high surface area effectively reduces the current density per unit surface area [51].

The rheological method is an advanced version of the solid-state method which overcomes the inhomogeneous or impure product obtained by the conventional solid-state reaction method and avoids careful control of the experimental conditions by wet chemistry [52, 53]. Yin et al. [52] reported that the initial discharge capacity of Li4Ti5O12 prepared by the modified rheological phase method was 176.7 mAh g−1 at 1 C, 170.4 mAh g−1 at 2.5 C, and 140.2 mAh g−1 at 10 C in the voltage range between 1.0 and 3.0 V, and the discharge capacities of 161.6, 156.5, and 112.3 mAh g−1 after 50 cycles at 1, 2.5, and 10 C, respectively. The high-rate performance is attributed to the pure and well-crystallized Li4Ti5O12 with ultrafine particles and narrow size distribution.

Sol–gel methods generally offer products with a homogeneous distribution of uniform, submicron size particles with good stoichiometric control. Several studies on LTO synthesis by the sol–gel method report the particles to have a uniform, nearly cubic structural morphology with narrow size distribution and exhibit high capacity and good cycling performance [20, 45, 54–59]. The sintering temperature also plays a major role which generally is optimized at 700–800 °C [60–64]. The influence of various chelating agents such as acetic acid (AA), citric acid (CA), oxalic acid (OA), and triethanolamine (TEA) used for the synthesis on the electrochemical property of LTO has been reported in a group of studies [65–67]. The LTO synthesized using TEA and CA exhibited discharge capacities of 150 and 137 mAh g−1, while the samples prepared by OA and AA delivered, respectively, 133 and 117 mAh g−1 after 30 charging/discharging cycles at a constant current density of 0.5 mA/cm2 [66]. During synthesis, chelating agents help the formation of homogeneous gel which after heat treatment leads to the formation of submicron LTO powders with narrow particle size distribution and superior dispersion [67].

Recent studies revealed that the rate capability of LTO can be improved by having more porosity and less agglomeration of the active materials by assembling them into 3D architecture and by careful control of the wall thickness of LTO hollow spheres [68, 69]. The studies showed that LTO grains of about 100 nm size get assembled on the walls of the hollow spheres, thereby reducing the Li diffusion distance and increasing the contact area between LTO and electrolyte. In addition, since more numbers of LTO primary grains are in contact with acetylene black, the electronic conductivity is improved.

Surfactants play an important role in the preparation of LTO with good dispersity. Pluronic P123 and cetrimmoniumbromide (CTAB) are the commonly used soft templates to prepare ordered mesoporous LTO particles [70–74]. In comparison to the densely aggregated LTO particles prepared without using P123 surfactant, the well-dispersed sample showed higher lithium storage capacity, especially at high discharge/charge current rates [70]. Feckl et al. [74] introduced a solvothermal reaction in tert-butanol in the absence of water for the synthesis of LTO. The results suggest that for the formation of phase pure LTO, the presence of a polymer (Pluronic P123) is needed. In addition, it affects the crystallization process by coordination to metal precursors, thereby modulating their reactivity. The morphological studies revealed the formation of highly crystalline material composed of spinel nanocrystals interconnected into a porous scaffold. Figure 6 shows the SEM and TEM images, respectively, of LTO prepared at 400 °C and 500 °C, by this route. The as-prepared nano-LTO spinel can be charged within a short time to almost its full capacity and then discharged at much slower rates, hence can find applications in electric and hybrid electric vehicles. Chen et al. [75] were able to synthesize mesoporous nest-like LTO particles of the size 8.2 nm using CTAB surfactant through a hydrothermal method. Figure 7 shows the SEM images of LTO synthesised using different concentrations of CTAB. The as-synthesized material offers low polarization with higher lithium diffusion coefficient and higher exchange current density.

Fig. 6
figure 6

SEM and TEM images, respectively, of nanosized spinel LTO heated at 400 °C (a, c) and 500 °C (b, d) (reproduced with permission from [74])

Full size image
Fig. 7
figure 7

SEM images of LTO synthesized with different concentrations of CTAB surfactant (a 0 M, b 0.1 M, c 0.2 M, d 0.5 M, the inset is TEM image of d) (reproduced with permission from [75])

Full size image

Hydrothermal synthesis on the other hand provides an efficient approach for the synthesis of nanostructured materials under milder conditions, which offers various morphologies and easy control [76–83]. From previous works, it is observed that an electrode with a nanoflower-like morphology will exhibit an excellent cycling performance and can be a good candidate as a next generation anode material for high-rate Li-ion batteries. Lin et al. [81] synthesized nanoflower-like LTO by means of a self-assembled synthesis route using amorphous TiO2 (400 nm) and LiOH as precursors. Under hydrothermal conditions, the sol–gel precursor is converted to layered hydrous LTO, which on calcination gets converted to nanoflower-type LTO (Fig. 8). As per their studies, the low thickness of LTO nanoplates (~10 nm) leads to the shortening of lithium transportation path, thereby exhibiting very high capacity. Cubic spinel LTO nanocrystals synthesized through a hydrothermal method via an intermediate (Li0.4H0.6)2TiO3 delivered excellent rate capability and cycling stability [82]. The synthesized material possesses a crystalline size of around 40 nm and is well dispersed. Figure 9 shows the TEM and SEM images, respectively, of (Li0.4H0.6)2TiO3 nanocrystals and LTO single crystals. The material delivers an initial discharge capacity of 184 mAh g−1 at 0.5 C. The Li+ ion diffusion path provided by the nanostructure accounts for this enhanced performance of the material. Yan et al. [84] reported that the spherical LTO with a particle size of 0.5 μm obtained through a hydrothermal method delivers a discharge capacity of 160 mAh g−1 at 0.2 C after 70 cycles and 123 mAh g−1 at 2 C after 150 cycles. LTO in the form of nanocables can also be synthesized by adopting a sol–gel-assisted hydrothermal method. Shen et al. [64] synthesized morphology-controlled co-axial nanocables consisting of multiwalled nanotube (MWNT) core and crystalline LTO sheaths (Fig. 10). The material delivered an initial discharge capacity of 171 mAh g−1 at 0.2 C and excellent rate capacity even at higher rates. The reason was ascribed to the improved Li+ ion and electron transport as a result of porous, nanosized morphology of LTO sheaths as well as the highly conducting MWNT cores.

Fig. 8
figure 8

SEM images of a amorphous TiO2 submicron spheres by the sol–gel method, b layered hydrous LTO after hydrothermal process, and c nanoflower-like spinel LTO after calcination (reproduced with permission from [81])

Full size image
Fig. 9
figure 9

TEM and SEM images, respectively, for (Li0.4H0.6)2TiO3 nanocrystals (a, c) and LTO single crystals (b, d) (reproduced with permission from [82])

Full size image
Fig. 10
figure 10

Typical FESEM images of MWNT@LTO coaxial nanocables; the arrows indicate the feature of the core/sheath nanocable structure (reproduced with permission from [64])

Full size image

The microwave-assisted synthesis method has distinct advantages over conventional methods in terms of lower synthesis temperature, rapid volumetric heating, higher reaction rate, shorter reaction time, and smaller particle size [85–89]. Yang et al. [88] reported the synthesis of Li4Ti5O12, Li2Ti3O7, Li2TiO3, and LiTiO2 from Li2CO3, TiO2, and Ti metal powder in different molar ratios through a hybrid microwave process. Microwave frequency of 2.45 GHz was used for irradiation. The synthesized Li4Ti5O12 showed an initial discharge capacity of 150 mAh g−1 and 94 % capacity retention even after the 27th cycle. By combining the advantages of both hydrothermal and microwave methods, Liu et al. [90] synthesized LTO with nanoflower-like and nanoparticle morphologies. The nanoflower-like morphology exhibited better reversibility than the nanoparticle LTO during the insertion/extraction process of Li+ ions into/from LTO.

Cellulose-assisted combustion synthesis technique has the advantages of both solid-state reaction and the sol–gel method, i.e., simple synthesis, low-temperature calcination, use of cheap raw materials, and good electrochemical performance of the product at high charge/discharge rates [66, 91–93]. Raja et al. [94] reported a novel aqueous combustion process using alanine as fuel for the synthesis of LTO powder with uniform morphology and an average particle size of 40–80 nm. The optical band gap was found to be 1.80 eV, in good agreement with the theoretical values. This study showed the presence of empty 3d bands of Ti in LTO, which gives an insulating nature to the material. Molten salt method and sonochemical methods are also efficient for the synthesis of LTO powders [95–97].

Spray pyrolysis is another useful method for the preparation of various types of electrode materials for Li-ion cells [39, 98–109]. Fine-sized LTO powder with spherical morphology has been synthesized through this method [105]. Figure 11 shows the SEM images of the post treated LTO powders obtained from the spray solutions with a) glycerol and b) formamide as drying control agents. Spherical LTO particles can be synthesized through an emulsion–gel process also [110]. The particle size of the synthesized material under the optimized condition is 0.45 μm. The suitability of heat treatment in an infrared furnace for the crystallization of amorphous metal oxide powders without aggregation is reported in this study. The microemulsion process is another effective way for the synthesis of LTO in which one can control particle size and size distribution. The LTO particles synthesized through this method by Liu et al. [111] exhibited the specific capacities of 173.7, 166.2, 136.2, and 130.2 mAh g−1 at 0.2, 1, 5, and 10 C after 50 cycles, respectively. Recently, Nugroho et al. [112] reported that hierarchical LTO microspheres can be prepared in super cooled methanol in a very short reaction time (less than 15 min) without utilizing structure directing chemicals. This type of an approach enables the researchers to control the primary particle size and mesoporosity of the active material.

Fig. 11
figure 11

SEM images of post treated LTO powders obtained by spray pyrolysis with different drying control agents: a glycerol and b formamide (reproduced with permission from [105])

Full size image

Theoretical studies and structure analyses on LTO

A few studies have reported the crystal structure of the electrochemically Li-ion-intercalated material Li4 + x Ti5O12 [113–117]. The Li-ion intercalation was considered to occur at the 16c site using the electrochemically Li-ion-intercalated samples [118–120], whereas a recent study reported it to be the three crystallographic sites 8a, 16c, and 48f, using the chemically Li-ion-intercalated polycrystalline samples [116]. By using the difference Fourier synthesis map, Kataoka et al. [121, 122] reported that the occupation sites for the intercalated Li-ions were the 8a and 16c sites. The studies on site population refinements showed that the chemical composition of Li-ion-intercalated Li4 + x Ti5O12 compound was Li5.35Ti5O12.

Structural and electronic properties of LTO were studied from density functional theory based first principles calculations which showed that the lattice constant shrinks very slightly after Li+ ions are intercalated into the lattice of LTO. The average intercalation voltage is calculated to be 1.45 V and the calculated electronic structures have shown that the delithiated state LTO is insulating while the lithiated state Li7Ti5O12 is metallic with partly filled Ti d-states [22].

The two-phase reaction between the end members LTO and Li7Ti5O12 results in the unusual combination of fast charge/discharge rates [123] and an extremely flat potential [115, 124, 125]. Recent studies showed that both these end members are described by the cubic space group Fd3̅m and have very similar lattice parameters 8.3595 Å and 8.3538 Å, respectively [115, 126]. As a result, the two phase reaction will not lead to substantial structural strain, the main factor which contributes to capacity loss in lithium battery electrodes. The solid-solution-induced disorder, resulting from the mixed 8a/16c occupation, is most likely responsible for the high-rate capabilities in Li4+xTi5O12[127]. The AC impedance studies and lattice gas model based Monte Carlo simulations showed the two phase co-existence state of Li-poor phase and Li-rich phase during the charge/discharge processes [128].

Li-ion insertion into the spinel structure occurs at a potential of approximately 1.5 V vs. Li/Li+ [129]. The electrochemical reaction can be described as:

$$ {\left[\mathrm{Li}\right]}_{8\mathrm{a}}{\left[\right]}_{16\mathrm{c}}{\left[{\mathrm{Li}}_{1/3}{\mathrm{Ti}}_{5/3}\right]}_{16\mathrm{d}}{\left[{\mathrm{O}}_4\right]}_{32\mathrm{e}}+{\mathrm{Li}}^{+}+{\mathrm{e}}^{-}\to {\left[\right]}_{8\mathrm{a}}{\left[{\mathrm{Li}}_2\right]}_{16\mathrm{c}}{\left[{\mathrm{Li}}_{1/3}{\mathrm{Ti}}_{5/3}\right]}_{16\mathrm{d}}{\left[{\mathrm{O}}_4\right]}_{32\mathrm{e}} $$

The spinel [Li]8a[Ti5/3Li1/3]16d[O4]32e structure provides a 3D interstitial space of face-shared 8a tetrahedral and 16c octahedral for lithium-ion transport through the crystalline bulk structure [130].

Electrochemical performance of LTO in Li-ion cells

Evaluation of LTO as anode material in combination with different cathode materials has been reported by various researchers. A few examples are discussed in this section. Figure 12 shows the comparison of operating voltage of cells with LTO anode when combined with various cathode materials. Since Li+ intercalation in LTO takes place at higher potential, the voltage of the cell with LTO as anode is less than that of graphite/carbon-based cells. This necessitates coupling of LTO with high voltage cathode materials to get appreciable operating voltage for the cell [23, 131–133].

Fig. 12
figure 12

Voltage of a Li-ion cell with Li4Ti5O12 anode and various cathode materials (reproduced with permission from [23]). LMO-LiMn2O4, LCO-LiCoO2, LNO-LiNiO2, L333-LiNi0.33Co0.33Mn0.33O2, LFP-LiFePO4

Full size image

LTO vs. LiCoO2

In this couple, the positive and negative electrodes are based on transition metal oxide electrodes, which have the capability of accommodating a significant amount of lithium within the host electrode structure. Lithium intercalation into LTO and deintercalation out of LiCoO2 is depicted in Fig. 13. By restricting to shallow limits of charge and discharge, the structural integrity of the electrodes can be maintained, and this permits high life cycle to be obtained with this couple [7].

Fig. 13
figure 13

Intercalation into the Li4Ti5O12 spinel and deintercalation out of the LiCoO2 layered structure form an ideal cell couple (reproduced with permission from [7])

Full size image

Figure 14 illustrates the voltage vs. the state of charge curve for LiCoO2–graphite and LiCoO2–LTO cells. In the case of LiCoO2–graphite, there is no end-of-charge indicator in the voltage profile that can signal the onset of catastrophic oxygen evolution from the LiCoO2. But the LTO–LiCoO2 couple shows a sharp rise in voltage around 2.75 V, indicating the end of charge of the cell. This demonstrates the inherent safety mechanism of the LTO–LiCoO2 cells [7, 134–136].

Fig. 14
figure 14

Inherent safety provided by the rise in voltage at the end-of-charge for the LTO–LiCoO2 cell couple (100 % SOC based on anode) (reproduced with permission from [7])

Full size image

LTO vs. Li1.1Ni0.3Co0.3Mn0.3O2(L333)

Lu et al. [137] reported a LTO/L333 cell, which exhibited excellent cycling performance for over 200 cycles at both ambient and elevated temperatures. The capacity retention of the LTO/L333 and MCMB (Meso Carbon Micro Beads)/L333 systems at 55 °C is about 96.5 and 83.5 %, respectively, after extensive cycling when subjected to the same test conditions.

LTO vs. LiFePO4

The couple, LTO vs. LiFePO4, is advantageous in terms of rate capability, cyclability, and safety [138, 139]. The disadvantage with this couple is its low voltage. Electrochemical results of a low-cost system with LTO/C vs. LiFePO4/C have been reported by Zaghib et al. [138]. The fast charge capability of the cells using ionic liquids based on 1 M LiTFSI in propylene carbonate and gamma butyrolactone was reported. After the first cycle, the coulombic efficiency was ~99 %, the average voltage was ~1.83 V, and continuous charge–discharge for 300 cycles showed good cyclability behavior (less than 1 % loss). This system yielded ~77 Wh kg−1 in a laboratory cell with both negative and positive electrodes using Al-expanded metal as current collector and a water-soluble binder. When the charging rates were 12 and 60 C, the cells showed 92 and 89 % of the nominal capacity, respectively.

LTO vs. LiMn2O4

Hu et al. [140] reported a hybrid battery–supercapacitor using a LTO anode and a LiMn2O4/activated carbon (AC) composite cathode. This system has advantages of both the high-rate capability from supercapacitor AC/LTO and the high capacity from secondary battery LiMn2O4/LTO. At 4 C rate, the capacity loss in constant current mode was <7.95 % after 5,000 cycles, and the capacity loss in constant current–constant voltage mode was <4.75 % after 2,500 cycles. The microbattery obtained through an interdigitated microarray of gold current collectors, coated with LiMn2O4/LTO and a gel–polymer electrolyte, was able to deliver a capacity of 300 mAh when operated at 2.5 V (Fig. 15) [141].

Fig. 15
figure 15

Schematic illustrations of a microarray electrodes of LiMn2O4 and LTO and b assembly of electrochemical cell (reproduced with permission from [141])

Full size image

LTO vs. LiNi0.5Mn1.5O4

LTO/LiNi0.5Mn1.5O4 cells showed that 83 % of initial capacity can be stored and delivered even after 1,100 cycles, indicating the excellent cyclability of the cells [142]. In addition, the cell with anode-limited design (N/P < 1) exhibited the best capacity retention, with only 2 % capacity loss after 50 cycles. Also, at room temperature, all cells exhibited excellent cycle stability, up to 60 cycles at 0.5 C charge and 2 C discharge rates [143].

LTO vs. Li2Co0.4Fe0.4Mn3.2O8

The Li-ion cell based on the combination of LTO anode with a high voltage-mixed spinel, Li2Co0.4Fe0.4Mn3.2O8 cathode, exhibited two well-defined plateaus, which correspond to the two-step Li intercalation in the mixed spinel cathode, as illustrated in Fig. 16. The results also confirmed the promising operational features of this cell which cycles with a high reversibility and good capacity delivery around an average of 3.5 V. However, for a cell voltage of 3.6 V, the potential at the Li2Co0.4Fe0.4Mn3.2O8 electrode was found to be around 5.2 V vs. Li, which is above the organic electrolyte stability limit [144].

Fig. 16
figure 16

Typical charge–discharge cycle of a LTO-LiPF6-PC-Li2Co0.4Fe0.4Mn3.2O8 cell (reproduced with permission from [144])

Full size image

LTO vs. Li–indium cell

LTO vs. lithium–indium (Li–In) cell developed by Kitaura et al. [145] using the 80Li2S · 20P2S5 (mol %) solid electrolyte retained the reversible capacity of about 90 mAh g−1 at 3.8 mA cm−2. In order to improve the rate performance, the lithium titanate particles were pulverized to decrease the particle size. The Li–In/70Li2S · 27P2S5 · 3P2O5 glass–ceramic/pulverized LTO cell was charged at a current density higher than 3.8 mA cm−2 and showed the reversible capacity of about 30 mAh g−1 even at 10 mA cm−2 [146].

Recent advances in LTO

Though there are many advantages associated with LTO, a major hurdle limiting its practical applications is its poor electrical conductivity (<10−13 S cm−1) [147]. In addition, the chemical diffusion coefficient is also low, about 2 × 10−8 cm2 s−1 at 25 °C [148, 149]. The poor electronic conductivity is attributed to the occupied 8a tetrahedral positions; the conductivity gets increased at higher temperatures due to the transfer of Li-ions from the 8a to vacant 16c octahedral positions [150]. To improve the conductivities, several effective ways have been proposed, including new synthesis methods; doping with metal or nonmetal ions in Li, Ti, or O sites; and incorporation of second phase with high electronic conductivity [53, 63, 76, 151–154]. This section deals with such advances in LTO.

Synthesizing a composite of LTO and a conductive second phase can improve the electronic conductivity and helps in controlling particle growth with a uniform coating of conductive phase on the surface of LTO [155–162]. The addition of carbon to the precursor material may increase interparticle contacts and hinder undesirable particle growth. A comparative study of the charge–discharge properties of pristine LTO and LTO/C made by Jung et al. [163] showed that LTO/C performed much better than pristine LTO. The commonly studied carbon sources are polyacrylic acid (PAA), polyacrylonitrile (PAN), starch, citric acid (CA), maleic acid (MA), and polyvinyl alcohol (PVA) [164–167]. The studies reported that the carbon source and carbon content influence the physical and electrochemical performances of LTO/C composites significantly [168–172]. He et al. [173] first reported the reduction reactivities of electrolyte on the uncoated and carbon-coated LTO electrodes. As per their studies, the carbon coating layers cover the catalytic active sites of LTO particles and separate the LTO particles from the electrolyte. Amorphous carbon-coated high grain boundary density dual-phase LTO-TiO2 nanocomposite anode material with good electrochemical performance in terms of high capacity (166 mAh g−1 at a current density of 0.5 C), good cycling stability, and excellent rate capability (110 mAh g−1 at a current density of 10 C up to 100 cycles) was reported [174]. Mesoporous, nanoarchitectured LTO/C is also reported to give high specific capacity and excellent cycling performance [175, 176]. The thickness of carbon coating also played a major role in ion diffusion. With increase in thickness, the apparent diffusion coefficient decreased, whereas the electronic conductivity increased [30]. However, the presence of too many additives significantly reduce the energy density of LTO anodes. Hsieh et al. [177] developed a new way to resolve this problem by the successful deposition of Zn layers on to spinel LTO crystals through a microwave-assisted reduction process. The results indicate that the as-synthesized Zn-LTO material delivers an initial discharge capacity of 183 mAh g−1 at 0.2 C rate. This can be ascribed to the fact that well-dispersed Zn layer offers an electronic pathway over the Li4Ti5O12 powder, thus imparting electronic conduction and reducing cell polarization.

The positive influence of carbon nanotubes (CNTs) on the electronic conductivity in LTO/CNT composite has been explored [178, 179]. The material yielded discharge capacities of 145 and 135 mAh g−1, respectively, at 5 and 10 C rates and could be cycled even at 20 C. Carbon nanofibers are also used for the same purpose [180, 181]. The reports suggest that the nanocomposite–LTO/CNF could overcome the inherent problems of LTO like poor Li+ diffusivity and poor electronic conductivity. Fang et al. [182] reported the synthesis of LTO/AB/MWCNTs through a solid-state reaction route. The material exhibited excellent high-rate performance and cycling performance, which was ascribed to the improved electroconductive behavior of the material. In place of CNTs, one can use graphitized carbon nanotubes (GCNTs) to improve the electrochemical performance [183]. The GCNTs provide a valid conductive network in the electrode system and control the particle size of LTO during synthesis (Fig. 17). The presence of GCNTs reduces the agglomeration during the synthesis. The initial discharge capacity offered by the material is 163 mAh g−1 at 0.5 C and 143 mAh g−1 at 10 C. LTO/GCNTs composite exhibits excellent rate capability compared to pristine LTO (Fig. 18). Recently, Shen et al. [184] introduced a new method for constructing high performance Li-ion batteries based on binder-free, flexible LTO/carbon textile composite. Carbon textiles improve the electrical conductivity of the material and serve as a current collector, thereby avoiding the use of conducting material and binder (Fig. 19). The material exhibited excellent rate capability and cyclic stability.

Fig. 17
figure 17

SEM images of the LTO/GCNTs composite (a) and the LTO (b) (reproduced with permission from [183])

Full size image
Fig. 18
figure 18

Rate capability of LTO/GCNTs (A) and LTO (B) (reproduced with permission from [183])

Full size image
Fig. 19
figure 19

Schematic drawing of a the conventional electrode and b novel binder-free flexible 3D electrode (reproduced with permission from [184])

Full size image

Another carbon-based strategy to improve the electrochemical performance of LTO is by synthesizing LTO/graphene nanocomposites [185]. In this composite, the electron transport is improved by forming a conductive graphene network throughout the insulating LTO nanoparticles. Reports suggest that such hybrid nanostructured anode material showed ultra high-rate capability and good cycling properties at high rates [186, 187]. Rai et al. [188] synthesized LTO/graphene nanocomposites through a solvothermal method which exhibits higher capacities and enhanced cycle performances within the voltage range of 1.0–2.5 V at a current rate of 10.4 C.

Surface modification using Al2O3 is another effective way to improve the performance of LTO. The as-prepared material exhibited improved cycling performance and high coulombic efficiency [189]. Zhang et al. [190] fabricated LTO/TiN nanocomposites through high-energy ball milling technique, which showed a high capacity of 130 mAh g−1 at a charge/discharge rate of 20 C. In addition, the material offered capacity retention of 85 % after 1,000 cycles at 10 C. A new approach to synthesize LTO with excellent capacity and cycling stability was introduced by Chen et al. [191] by means of developing defective mesoporous Li4Ti5O12 − y via a solvothermal method. The material delivered a discharge capacity of 139 mAh g−1 at a high rate of 20 C with a capacity retention of 91.4 % over 300 cycles. The excellent performance was attributed to the mesoporous structure with oxygen vacancies and Ti3+–O2−–Ti4+ pairs which enable improved electronic conductivity to the material.

The effect of doping on the electrochemical performance of LTO has been reported in a group of studies [192–221]. The improvement in performance has been attributed to the effects of the dopant on the LTO microstructure [23, 192]. Doping improves the diffusion coefficient and cycle life of LTO. Doping with Cr, Al, and Nb ions considerably improved the reversible capacity and cycling stability of LTO [192–197]. Gao et al. [198] reported the synthesis of a spherical LTO/C and La3+-doped LTO/C composites via an “outergel” method using carbon black. Sha et al. [199] reported a new concept for developing Li+-ion conducting lithium lanthanum titanate solid electrolyte and LTO composite through one-pot combustion technique. The results show that the sample exhibits a relatively high capacity of 113.5 mAh g−1 even at a discharge rate of 40 C. Even though Mn doping can increase the initial capacity of LTO, its cycling performance gets retarded. Wang et al. [200] reported Ru-doped nanostructured anode materials synthesized through a reverse microemulsion method. The results show that because of the bigger ionic radius of Ru+ than Ti4+, the lattice constant of the doped materials is enlarged, which is in favor of the insertion and extraction of lithium ions. The as-prepared material exhibited a specific discharge capacity of 131 mAh g−1 over 100 cycles even cycled at a rate of 60 C. Doping with Mg, Sc, and Cu ions improved the conductivity of LTO [201–207], whereas with Al, the specific capacity of LTO was decreased but its cyclability at high charge/discharge rate was improved [208, 209]. Studies showed that the spinel structure of LTO was more stable when Ti4+ was substituted by Al3+ [210, 211]. Vanadium-doped LTO has lower electrode polarization and high Li-ion diffusivity in solid state, which showed that V doping is beneficial to the reversible intercalation and deintercalation of Li+ [212, 213].

Studies were also carried out for the development of carbon-coated, doped LTO porous particles. Zhang et al. [222] synthesized carbon-coated Li3.9Sn0.1Ti5O12 through a new sol–gel process in which a thin layer of amorphous carbon is coated on the doped particles that contain numerous nanopores. They suggested that the synthesized material can exhibit almost 100 % capacity retention after 100 cycles of charge–discharge at 5 C. Studies on Li4Ru0.01Ti4.99O12/C reports a specific capacity of 120 and 110 mAh g−1 at 5 and 10 C rate after 100 charge/discharge cycles [223].

Wu et al. [224] investigated the effect of current collector on the charge/discharge capacity and cycle stability of LTO electrode at high C rates. As per their studies, in order to achieve high power capability of an electrode material, the overall resistance of the electrochemical system should be minimized. They used different types of current collectors, including Al foil, etched Al (E-Al), etched Al with a C coating (C-E-Al), Cu foil (Cu), and Cu foil with a C coating (C-Cu). The results suggest that the capacities of LTO electrodes above 1 C rate are in the order of Al < E-Al < Cu ~ C-E-Al < C-Cu, exhibiting remarkable enhancement in rate performance by the C coating for both metals.

Nanolithium titanate

The use of nanostructured electrode materials showed improved properties in terms of energy density, rate capability, and cycle life compared to micrometer-sized materials [40–42]. The studies that reported on the synthesis and characterization of nano-LTO [46, 225, 226] show that nanosizing is an effective way to overcome poor Li+ diffusion coefficient of LTO. The 1D LTO nanowires synthesized by Kim et al. [227] delivered an initial capacity of 165 mAh g−1 at 0.1 C with a capacity retention of 93 % even at 10 C, attributed to the more active surface in the nanowires providing shorter pathways for lithium diffusion. Similarly, the 2D LTO nanosheets with cubic spinel structure synthesized via a lithiation process of exfoliated titanate 2D nanosheets and a subsequent heat treatment at an elevated temperature of 600 °C show promising discharge capacity with excellent cyclability [228].

Electrospinning is nowadays attempted as a versatile method for the preparation of nanosized architectures of electrode-active materials for Li-ion battery. The fibrous morphology with good porosity improves the conductivity of the electrode and ion mobility. The synthesis of LTO fibers by electrospinning and their electrochemical performance evaluation has been reported in a few studies [21, 229, 230]. Polyvinylpyrrolidone (PVP) is the polymeric support generally employed for the preparation of precursor for electrospinning. The electrospun LTO material exhibited stability in 3D network architectures and good reversibility in electrochemical performance [230]. Recently, in our laboratory, LTO fibers were synthesized by electrospinning of the precursor solution containing lithium acetate, titanium isopropoxide, PVP, and acetic acid. The SEM image of the product obtained after calcination of the electrospun precursor at 800 °C for 4 h showed the formation of LTO in ultrafine fiber form, with fiber diameter in the range of 400 to 700 nm (Fig. 20) [21]. The well-crystallized graphene-embedded LTO nanocomposites obtained after calcination of the electrospun precursor at 550 °C in N2 atmosphere showed ~91 % retention of the initial capacity even after 1,300 cycles at 22 C, as reported by Zhu et al. [231]. Guo et al. [232] made a comparative study for the electrochemical performances of LTO/C fibers and LTO/C particle/fibers. Among the two electrospun composites, LTO/C fibers exhibited better kinetics than LTO/C particles/fibers as a result of the elimination of LTO aggregates and the formation of carbon-based fiber structure. Wang et al. [233] synthesized mesoporous LTO/carbon nanofibers via electrospinning of the precursor gel at very high voltage of 13.5 kV. The as-prepared material exhibited an initial discharge capacity of 127.4 mAh g−1 at a high rate of 5 C. According to them, the excellent electrochemical performance of the material may be ascribed to the mesoporous structure inside the nanofibers, which enables the easy infiltration of the electrolyte into the mesopores to contact the LTO nanocrystals. In addition, the 1D nanostructure composed of nanosized LTO and carbon layer may improve the kinetics of Li+ intercalation.

Fig. 20
figure 20

SEM images of LTO fibers (reproduced with permission from [21])

Full size image

Based on the above discussion, a comparative data of LTO with different morphologies giving excellent electrochemical performances have been compiled as shown in Table 1. From the results, it is well understood that the morphology of the material plays a significant role in the electrochemical performance. LTO in the form of mesoporous structures, nanoflowers, multiwalled nanotubes, etc. shows excellent electrochemical performances. Such materials exhibited high-rate capability with good cycle stability.

Table 1 Comparative data of LTO with different morphologies giving excellent electrochemical performance
Full size table

Lithium titanate in commercial batteries

Nanolithium titanate is used as anode material in “Nanosafe” batteries produced by M/s. Altairnano (finance.yahoo.com/…/altairnano-lithium-titanate-energy- storage-150000131.html). The battery has a specific power of 4,000 W kg−1 and retains 85 % of capacity after 20,000 cycles. LTO is also used in Super Charge Ion Battery (SCIB) manufactured by M/s. Toshiba (http://en.wikipedia.org/wiki/Lithium%E2%80%93titanate_battery). These batteries possess an energy density of 89 Wh kg−1 and are designed to recharge to 90 % capacity within 10 min. The SCIBs can support >6,000 cycles and are safer than conventional Li-ion batteries. The batteries can be operated over a full range of state of charge (SOC) from 0 to 100 %, whereas most other lithium chemistries limit SOC from 30 to 80 %. Recently, in order to decrease the price of Li+-ion batteries as well as to reduce the technology’s environmental burden, an aqueous fabrication process of LTO electrode using Acryl S020 as binder was established [234]. The electrode fabricated showed relatively good cycling stability and high capacity retention.

Summary

Being a zero strain material, the use of LTO as anode imparts good cycle life and rate capability to the Li-ion battery and enhances the safety features as well. The performance of LTO could be further improved when it is used in nanostructured forms. Doping with suitable anions or cations can improve the electrical conductivity of the material, thereby improving the power capability of LTO. Carbon coating is another efficient method to improve the electrical conductivity of the material; this helps to control the particle size of LTO also. The Li-ion cells with LTO anode having the favorable characteristics for automotive applications are being explored intensively.