Recent Progress of TiO 2-Based Anodes for Li Ion Batteries

TiO 2 -based materials have been widely studied in the field of photocatalysis, sensors, and solar cells. Besides that, TiO 2 -based materials are of great interest for energy storage and conversion devices, in particular rechargeable lithium ion batteries (LIBs). TiO 2 has significant advantage due to its low volume change (<4%) during Li ion insertion/desertions process, short paths for fast lithium ion diffusion, and large exposed surface offering more lithium insertion channels. However, the relatively low theoretical capacity and electrical conductivity of TiO 2 greatly hampered its practical application. Various strategies have been developed to solve these problems, such as designing different nanostructured TiO 2 to improve electronic conductivity, coating or combining TiO 2 with carbonaceous materials, incorporating metal oxides to enhance its capacity, and doping with cationic or anionic dopants to form more open channels and active sites for Li ion transport. This review is devoted to the recent progress in enhancing the LIBs performance of TiO 2 with various synthetic strategies and architectures control. Based on the lithium storage mechanism, we will also bring forward the existing challenges for future exploitation and development of TiO 2 -based anodes in energy storage, which would guide the development for rationally and efficiently designing more efficient TiO 2 -based LIBs anodes.


Introduction
Lithium ion batteries (LIBs) are becoming the best choice in portable electronics, implantable devices, power tools, and hybrid/full electric vehicles (EVs) for their high working voltage, low self-discharge rate, long cycle life, high energy, and power density [1,2].Using electric vehicles instead of traditional gasoline powered transportation can significantly reduce pollution of combustion gas and increase energy security.More importantly, the high energy efficiency of LIBs also has potential application in various large electric grid applications, including improving the energy efficiency of wind, solar, tidal, and other clean energy; thus LIBs are expected to have a very favorable impact on building an energy-sustainable economy [3,4].Figure 1 shows the forecasted evolution of the LIBs demand in the future years [5]; we think we will see economical battery-driven electric vehicles sooner than most people expect.
Up to now, the vast majority of commercial LIBs rely, at the cathode side, on transition metals oxides or phosphates active material (LiCoO 2 [6], LiNiO 2 [7], LiMnO 2 [8], LiFePO 4 [9], LiMnPO 4 [10], etc.), while graphite is commonly used as anode active material.Figure 2 is the principle of a typical lithium ion battery; both anodes and cathodes could shuttle lithium ion back and forth between them.The electrolyte is usually made of polypropylene/polyethylene which contains lithium salts (i.e., LiPF 6 ) in alkyl organic carbonates.The separator between anode and cathode can allow the diffusion of Li ions from cathode to anode during the charging and the reverse discharging process.
The anode is a crucial part in LIBs; therefore, the research and the development on the current situation of anode materials are one of the most important factors to determine the performance of this device.An ideal anode material shall meet the following requirements [11,12]: (1) high specific surface area and large exposed surface offering more lithium insertion channels, (2) low volume change during Li ion insertion/desertions process, which is important for good cycling stability, (3) large pore size and short paths for fast lithium ion diffusion with high speed, which is crucial for good rate capability, (4) low internal resistance which leads to fast charging and discharging, (5) low intercalation potential for Li, (6) low price, (7) environment friendly.Based on the  Reprinted from [5].
Li ion storage mechanisms, anode materials can be classified into the following categories: carbon based materials, such as graphite, amorphous carbon, carbon nanotubes, and graphene; alloy/dealloy materials, such as Si, Sn, Ge, Al, and Bi; transition metal oxides (M x O y , M = Cu, Mn, Fe, Co, Ni, etc.); metal sulphides; metal phosphides and metal nitrides [13][14][15].Figure 3 shows the potential versus Li/Li + and the corresponding capacity density of some potential active anode materials.In the whole, transition metal oxides always have relatively higher potential and capacity.Among these transition metal oxides, TiO 2 is one of the most promising anode candidates for LIBs, which exhibits excellent structural stability, high discharge voltage plateau (more than 1.7 V versus Li + /Li), excellent cycling stability, environmentally friendly, high safety, and low cost [16,17].However, some limitations of TiO 2 exist as well, such as low capacity, low electrical conductivity, and poor rate capability.Table 1 compares advantages and limitations of TiO 2 and other anode materials.The reversible lithium ion insertion and extraction from TiO 2 occur according to the following reaction [18]: where  can range between 0 and 1, depending strongly on the TiO 2 polymorph, particle size, and morphology.Therefore, the electrochemical performance of TiO 2 highly depends on their structural parameters such as crystallinity, size, morphology, polymorphs, and specific surface area.Table 2 summarizes the structural and electrochemical profiles of various TiO 2 polymorphs [19].Amongst these, the anatase, rutile, brookite, and bronze phases of TiO 2 have been reported for LIBs applications.However, there are some problems which exist in practical application, that is, low electrical conductivity (10 −12 -10 −7 s cm −1 ) and diffusion coefficient of lithium ions (10 −15 -10 −9 cm 2 s −1 ), always leading to the poor rate capability of TiO 2 anodes, which result from their low electric conductivity with the lack of open channels [20][21][22].
Based on the analysis of shortcomings of TiO 2 anodes, several different strategies have been developed to address these issues of TiO 2 -based anodes and summarized in this review, such as designing different nanostructured TiO 2 , coating or combining TiO 2 with carbonaceous materials and metal oxides to change the physical and chemical surface, and selective doping with heteroatoms to form more open channels and active sites for Li ion transport, as well as increasing the intrinsic conductivity.Indeed, these methods lead to many advantages in improving the capacity, cycling stability, and rate capability of TiO 2 .nanotubes, nanofibers, and nanowires could serve as an electron express way along the axial direction for electron collection due to a shorter collection time for the efficient electron transportation [31,32].For example, single-crystalline TiO 2 nanowires have an electron mobility of ∼1 cm 2 V −1 s −1 , nearly 1-2 orders higher than that of polycrystalline nanoparticles [33,34].Thus, 1D nanostructure is conductive to shorten the diffusion length for electrons and lithium, increase the electrode/electrolyte interfacial area, and accommodate volume changes arising from the lithium ion insertion/extraction process [35].Moreover, due to the unique structural flexibility, 1D material with good mechanical properties has potential in various binder-free and flexible electronics and photonics [36][37][38].1D TiO 2 with different nanostructure (Figure 4) including nanotubes, nanofibers, and nanorods has been designed for high performance anodes in LIBs.The significance of 1D TiO 2 on battery performance was demonstrated by several  groups.Tammawat and Meethong reported that anatase TiO 2 nanofiber anodes were directly used as an anode active material in LIBs without an additive or a binder.The nanofibers exhibited a high lithium storage capacity, a stable cycle life, and good rate capability [39].The enhanced reversible capacity and cycling performance of the anatase TiO 2 nanofibers are attributed to the large surface area of the nanofibers, small nanocrystalline size, large Li nonstoichiometric parameters, and the increased electronic conductivity.Armstrong et al. prepared TiO 2 nanowires; these unique structures gave a higher capacity of 305 mA h g −1 compared to 240 mA h g −1 of bulk TiO 2 [40][41][42].The enhanced capacity closely related to the good electronic conductivity and large surface area.For TiO 2 , 2D structures could provide stable framework, effective grain boundaries, and short path for lithium ion diffusion and storage compared with 0D nanoparticles and 1D nanostructures.Significant efforts have been made on the fabrication of 2D TiO 2 materials.Li et al. synthesized mesoporous TiO 2 nanoflakes with size of 10-20 nm via hydrothermal methods using Ti(SO 4 ) 2 as titanium source and NaOH solution as alkaline medium.The result of electrochemical performance test shows that the prepared TiO 2 nanoflakes with shorter calcining time have high discharge specific capacity (261.5 mA h g −1 ) and good cycling performance [25].In the process of heat treatment, longer calcining time results in uneven nanometer size and obvious reunion phenomenon.Shorter calcining time usually leads to more stable structure and higher specific surface area.Thus, both the lithium storage specific capacity of TiO 2 and the cycling stability of the battery can be improved [25].Zhu et al. first synthesized the mesoporous single-grain layer anatase TiO 2 nanosheets using a simple and easily reproducible method.The obtained TiO 2 nanosheets exhibited a discharge capacity of 73 mA h g −1 with obvious voltage plateaus over 4000 cycles, highlighting them as promising anode material for long-term LIBs [58].Wu et al. demonstrated a simple and green approach for the synthesis of anatase petal-like TiO 2 nanosheets; the unique structure showed high capacity and good cycling stability.This is because obtained petal-like TiO 2 nanosheets showed a comparative surface area of 28.4 m 2 g −1 , which should provide shorter diffusion distance for Li ions and should be beneficial for electrochemical performance of the electrode [29].Some typical TiO 2 nanosheets used in the lithium storage were listed in Table 3.It can be seen that 2D TiO 2 nanosheets exhibit the superior capacities, improved cycling stability and rate capabilities, owing to unique exposed facets, shortened path, and reserved porous structures.

Three-Dimensional Porous Structure.
Recently, threedimensional (3D) porous structure materials exhibiting interesting electrochemical performance in LIBs have attracted more attention, due to their special nature including highly exposed skeleton, tunable pore size, high porosity, high specific surface area, and low bulk density [59,60].As described, first, the unique structure is conductive to enhance the diffusion kinetics for its short diffusion paths for Li ions.Second, the pores are beneficial to enable easy infiltration of electrolyte and fast liquid-phase Li ion diffusion, reducing the concentration polarization and increasing rate performance and capacity of the cell.Third, the continuous network of 3D porous structure can provide better electrical conductivity  ∼135 mA h g −1 3400 mA g −1 compared to loosely connected particles.Forth, the porosity in 3D structure should help in accommodating volume change during charging/discharging process and maintaining the structural integrity of the electrode [61,62].Up to now, different hollow structures such as hollow spheres, nanoboxes, and nanotubes are explored to be used as high performance LIBs anodes [63][64][65][66].And the same happens for TiO 2 ; the introduction of porosity into TiO 2 nanomaterials also can improve the cycling stability and increase the capacity at high charge-discharge rates due to the increased contact surface area and shortened path length for diffusion of Li ions [67][68][69][70].Highly crystalline, nonordered mesoporous TiO 2 nanocrystalline with high specific surface area and having anatase as the dominant phase have been reported by Gerbaldi and other researchers, which showed very high rate capability and excellent stability upon very prolonged cycling [71,72].Besides, the storage characteristics of the mesoporous samples in lithium test cells were reported, and a close correspondence between the structural properties of materials and the electrochemical performance was studied.
The presence of mesopores is thought to be important for high rate performances and favorable for electrolyte ions transport.Lou's group recently reported the TiO 2 hollow spheres and submicroboxes, owing to the high surface area, porous shells, and small primary nanoparticles; these TiO 2 hollow structures possess significantly improved lithium storage properties with superior lithium storage properties in terms of high specific capacity, long-term cycling stability, and excellent rate capability [73,74].Figure 5 shows the typical morphology of different hollow TiO 2 structures, all of which exhibit outstanding electrochemical performance.

Carbon Coating.
Carbon coating is an effective and common approach to improve the electrochemical performance of the anode materials.The role of carbon has also been studied, such as reducing the charge transfer resistance and improving the Li ions diffusion, enhancing electron transport, buffering the large volume changes during the charge/discharge process, and acting as a passivation layer to prevent the aggregation of active materials [66,85,86].Some research has proved that the SEI (solid electrolyte interphase) film for carbon coated materials was found to be much thinner than the SEI film on uncoated active materials; thus, initial charge-discharge efficiency can be greatly enhanced [87,88].For example, Xia et al. investigated the effect of carbon coating on TiO 2 ; these TiO 2 /carbon hybrids could enhance electronic conductivity and provide flexible space for suppressing the large volume expansion during cycling [89].E. Portenkirchner reports that the anatase TiO 2− -C nanotubes demonstrate a superior Li storage capacity as high as 320 (±68) mA h g −1 , nearly twice as high as pure TiO 2− .Electrochemical impedance spectroscopy reveals smaller charge transfer resistances for TiO 2− -C nanotubes at the solid/liquid interface which improves the transfer of lithium ions from the electrolyte into the electrode [90].Besides, the composites also showed higher initial charge-discharge efficiency compared to pure TiO 2 ; the reason can be ascribed to the formation of thinner SEI films.Zheng and coworkers prepared nitrogen-containing carbon modified porous TiO 2 composites.The as-prepared composites also exhibited enhanced rate performance and superior cyclability for LIBs compared to pure TiO 2 (Figure 6).The study indicates that N doping is favorable to improve the electronic conductivity and the composites possessed much lower charge transfer resistance than that of TiO 2 [91].

Combining TiO 2 with Carbon Nanotubes (CNTs).
In recent years, CNTs have been approved to be a good anode material for lithium batteries due to their unique 1D structure, high conductivity (10 6 S m −1 for single-walled carbon nanotubes and >10 5 S m −1 for multiwalled carbon nanotubes), low gravity (0.8-1.8 g cm −3 ), high mechanical properties (Young's modulus of the order of 1.2 TPa), and high surface area (>100 m 2 g −1 ) [92][93][94][95][96]. Some studies showed that CNTs could exhibit reversible capacities anywhere from 300 to 1000 mA h g −1 after chemical treatment; the value is significantly higher than the theoretical capacity of graphite (320 mA h g −1 ) [97][98][99][100].Numerous CNTs conjugated with a variety of nanostructured materials and metal oxides have been synthesized to obtain good electrochemical performance [101][102][103].For example, CNTs@TiO 2 composites have been synthesized by controlled hydrolysis of titanium isopropoxide over CNTs (as shown in Figures 7(a) and 7(b)).When CNTs are used as lithium ion battery electrodes, their inclusion is beneficial for an extreme enhancement of the rate capability of lithium ion uptake and release in TiO 2 ; it also favors the interfacial lithium ion intake from the solution by reducing the inherent charge transfer resistance.CNTs efficiently provide electrons to the nanostructure through the formation of Ti-C bonds, then effectively assisting lithium ion incorporation [104].Zhao's group synthesized TiO 2 /CNTs composite through chemical vapor deposition method.The in situ synthesized composite showed better electrochemical performance (high specific capacity and long-term cycling stability) than the pristine TiO 2 .This is because CNTs in the composites not only supply an efficient conductive network but also keep the structural stability of the TiO 2 particles, ultimately resulting in the improved electrochemical performance [105].

Combining TiO 2 with Graphene.
Graphene is a single atomic plane of graphite and consists in a honey comb network of sp 2 carbons bonded into two-dimensional sheets with nanometer thickness, due to its unique properties, including high intrinsic carrier mobility (200 000 cm 2 V −1 s −1 ), relevant mechanical strength, excellent conductivity (5000 W m −1 K −1 ), high optical transmittance (∼97.7%),large theoretical specific surface area (2630 m 2 g −1 ), and superior mechanical strength which make graphene a suitable anode material for LIBs [105][106][107][108][109][110][111].Besides, the rich functional groups on the surface of graphene make it an appealing 2D substrate for the anisotropic growth of different kinds of active materials [112,113].For example, Fang el al. 's group prepared novel mesoporous graphene nanosheets with an excellent reversible capacity of 833 mA h g −1 after 60 cycles [114]; this capacity is much higher than the theoretical lithium storage of graphite.This can be ascribed to the high contact surface area for lithium ion adsorption and intercalation, as well as edges and other defects.Thus, many synthetic strategies have been reported for TiO 2 /rGO hybrid nanostructures; Ti-C bond in the hybrids is crucial for rapid interfacial charge transferring.Etacheri et al. chemically bonded mesoporous TiO 2 nanosheets to rGO sheets through a photocatalytic reduction method, resulting in the formation of Ti 3+ -C bonds between TiO 2 and rGO.These TiO 2 /rGO hybrid nanostructures demonstrate superior specific capacity, excellent rate capability, and capacity retention compared to a physical mixture of TiO 2 and rGO [115].
The reason can be attributed to the higher electrochemical performance of TiO 2 /rGO hybrid nanostructures to efficient interfacial charge transfer between TiO 2 nanosheets and rGO, which is fostered by Ti 3+ -C bonds.Figure 8 shows the SEM and digital images of TiO 2 /rGO hybrid films; insets in Figure 8(b) display the flexibility of the corresponding films upon bending.The high flexibility of graphene could be an excellent supporting matrix or coating layer to accommodate the volume change during the charge/discharge process.This is crucial for maintaining the good cycling stability [116].

Combining Metal
Oxides with TiO 2 .Combining different physical and electrochemical properties of components with TiO 2 and utilizing the respective advantage to increase the capacitance are a feasible method, such as using high conductive materials (conducting polymers) [117,118], increasing the surface area (carbon nanotubes) [104,105], and using high performance redox-active transition metal oxides (MnO 2 ) [119].Among the above materials, metal oxide coatings can efficiently improve the capacitive performance of the materials through intruding synergistic effects into an electrode system, such as in SnO x @TiO 2 core-shell composites, due to the nearly zero volume change of TiO 2 in insertion of Li + ions process, making it suitable as a backbone or protective layer for SnO x to restrain the pulverization and achieve an excellent high rate cycling ability and good cycling stability [120][121][122][123].
Recently, synergistic TiO 2 -MoO 3 core-shell nanowire arrays were prepared via a facile hydrothermal growth of ordered TiO 2 nanowires followed by a subsequent controllable electrodeposition of nano-MoO 3 .The composites exhibited high gravimetric capacity, good rate performance, and cycling stability., B [134,135], and N [136], all of which show beneficial effect on increasing the electrical conductivity more or less.
2.4.1.Ion Dopants.Doping Ti 3+ in the TiO 2 structure can provide conduction band electrons and undoubtedly improves its conductivity, which also helps to increase the reversible capacity.Ren et al. presented a simple and controllable method to prepare the Ti 3+ doped TiO 2 by a solvothermal process at lower temperature.The doped TiO 2 nanoparticles showed much enhanced electrochemical performance in reversible capacity, rate performance, and stability comparing with the pure TiO 2 [132].This is because Ti 3+ doping can increase the electrical conductivity of TiO 2 .Liu et al. synthesized Ti 3+ doped TiO 2 nanotube arrays which also exhibited excellent lithium ion intercalation performance with an initial discharge capacity of 101 mA h g −1 at a high current density of 10 A g −1 [137].The much improved lithium ion intercalation properties were attributed to the easy phase transition promoted by the surface defects, that is, Ti-C, Ti 3+ , and O 2− vacancies, which could serve as nucleation centers.In addition, the rate performance was also improved due to the enhanced electrical conductivity.Sn 4+ , Fe 3+ , and other metal ions were also investigated as dopants to improve the electrochemical performance [133,138].Kyeremateng and coworkers reported that the Sn doped TiO 2 nanotubes delivered much higher capacity values compared to simple TiO 2 nanotubes.The outstanding electrochemical behaviour is proposed to be related to the enhanced lithium diffusivity evidenced with Cottrell plots (Figure 10) and the rutile-type structure imparted with the Sn doping.The results showed that lithium ion insertion into Sn doped TiO 2 is about 40 times faster than into undoped TiO 2 [133].

Atom Dopants.
Atoms doping is also a useful technique to increase the internal surface area and electrical conductivity of anode materials.For example, boron (B) and nitrogen (N) doping had been proven to be an effective strategy for improving the electrochemical performance of carbon materials [139][140][141][142][143]. For example, B doped graphite has a larger lattice constant value,  0 , and a smaller  002 distance than ideal graphite, due to replacement of the carbon atoms with boron [144], leading to increases in both the crystallinity and electronic property of carbon as a Li-host material.nondoped TiO 2 sample [134].Furthermore, nitrogen doping has been proven to be an effective strategy for improving the capacity of TiO 2 .This is because nitrogen doping can improve the electric conductivity as well as the ionic conductivity; after introducing the N atoms, the distortion of Ti-O lattice can affect the electrochemical reactions on the interfaces between electrodes and the electrolyte, as well as lithium ion diffusion in the Ti-O lattice [68,[145][146][147].

Conclusions and Outlook
In summary, this review showed the amount of research efforts towards the development and improvement of TiO 2based anode materials for LIBs.Several elegant strategies aiming to boost the electrochemical performance and promote the practical application of TiO 2 have offered, including fabrication of nanostructures with different morphologies and sizes, modification by various coating materials (carbon materials and metal oxides), elements doping.The unique design allows achieving high lithium storage and good cycling stability based on the high lithium ion flux at the electrode/electrolyte interface, low internal resistance, short paths for fast lithium ion diffusion, and low volume change during Li ion insertion/desertions process.When combining these exquisite features together, it is possible for maximizing their electrochemical advantages to meet the present energy demands.
Firstly, the performance of TiO 2 depends strongly on its particle size and morphology.Therefore, different structures of TiO 2 are explored to improve the electrochemical performance of TiO 2 .In a second category, combining TiO 2 with carbonaceous materials such as active carbon, CNTs, and graphene, the composite anode materials can obtain moderate conductivity, large surface area, good chemical stability and mechanical property.In the third, metal oxides such as Fe 2 O 3 , SnO 2 , and MnO 2 can provide larger capacities and high energy density compared to pure TiO 2 , which had been combined to improve the overall anode performance.Fourthly, for the low electrical conductivity and ion diffusivity of TiO 2 , doping with appropriate ions or atoms is advantageous since this method can improve the intrinsic nature of TiO 2 by adjusting its electronic structure and forming more open channels and active sites for Li ion transport via the expanding interplanar spacing of TiO 2 lattices.
Finally, from this short review, we can conclude that high energy density, high cycle life, and high efficiency battery will still be the mainstream in the future growth of lithium batteries.In order to utilize the TiO 2 -based materials as effective anodes in commercial LIBs, interdisciplinary effort in this area is however required.
Although considerable advances have been achieved in improving the Li ion storage performance of TiO 2 , several fundamental issues are still needed to be solved.For example, nanomaterials usually show large surface area, which leads to more significant side reactions and results in low coulombic efficiency.Besides, nanopowder has lower density compared to the block material, which would reduce the volumetric energy density of battery.The following two possible strategies may be helpful to solve the abovementioned problem: (1) adopting surface modification or coating to reduce unnecessary side reactions; (2) designing hierarchical structures to enhance the tap density of anode materials.

Figure 2 :
Figure 2: Schematic representation of lithium insertion/deinsertion mechanism for current rechargeable lithium battery.

Figure 3 :
Figure 3: Comparing of some potential anode materials for lithium ion batteries.

Figure 6 :
Figure 6: The morphology of carbon coated TiO 2 spheres and their cycling performance.Reprinted from [91].

Figure 10 :
Figure 10: Cottrell plots for the determination of Li + diffusion coefficients in TiO 2 and Sn doped TiO 2 .Reprinted from [133].
[43]et al.reported a highly ordered anodic TiO 2 nanotube arrays with a tube length of 9 mm.These nanofibers exhibited significantly better microbattery performance (i.e., areal capacities, rate capability, and cycling stability) than both previously TiO 2 -based electrodes and other 3D microbattery electrodes.They suggested that the enhanced performance depends strongly on the long range ordering and crystallinity of the nanotube structures[18].Wang et al. prepared a hybrid Li ion capacitor based on TiO 2 nanobelt array and graphene hydrogels cathode.It is found that the densities of the capacitor can reach an energy density of 21 W h kg −1 and a high power density of 19 kW kg −1[43].The above studies also show that self-ordered 1D nanoarchitectures grown directly on a current collector are helpful to have a regularly oriented property and good contact with the current collector, enhancing the lithium ionic and electrical conductivities.Designing 1D structure is an effective way to improve the Li storage properties of TiO 2 .2.1.2.Two-DimensionalStructure. Two-dimensional (2D) nanomaterials often have large exposed surfaces and specific facets, which is very effective in high energy storage applications such as LIBs and supercapacitors.More importantly, 2D nanostructures can offer short ion diffusion length and open charge transport channel for electrolyte penetration and buffer the

Table 3 :
The capacity of reported 2D TiO 2 materials for lithium storage.