Single‐Crystalline TiO2(B) Nanobelts with Unusual Large Exposed {100} Facets and Enhanced Li‐Storage Capacity

The {100} facet of single‐crystalline TiO2(B) is an ideal platform for inserting Li ions, but it is hard to be obtained due to its high surface energy. Here, the single‐crystalline TiO2(B) nanobelts from H2Ti3O7 with nearly 70% {100} facets exposed are synthesized, which significantly enhances Li‐storage capacity. The first‐principle calculations demonstrate an ab in‐plane 2D diffusion through the exposed {100} facets. As a consequence, the nanobelts can significantly accommodate Li ions in LiTiO2 formula with specific capacity up to 335 mAh g−1, which is in good agreement with the electrochemical characterizations. Coating with conductive and protective poly(3,4‐ethylenedioxythiophene)‐poly(styrenesulfonate), the cut‐off discharge voltage is as low as 0.5 V, leading to a capacity of 160.7 mAh g−1 after 1500 cycles with a retention rate of 66% at 1C. This work provides a practical strategy to increase the Li‐ion capacity and cycle stability by tailoring the crystal orientation and nanostructures.


Introduction
TiO 2 has been a widely investigated anode material for Li-ion battery because of its low cost and high stability. In particular, TiO 2 (B) [1] is one of typical battery materials that exhibits pseudocapacitive behavior and thus interesting for highpower applications. It has a higher capacity and better overcharge protection as compared with graphite. TiO 2 (B) can theoretically reach a high capacity to the formula Li 1. 25 TiO 2 (419 mAh g −1 ) [2] with a high redox potential of ≈1.5 V versus Li/Li + . So far, the TiO 2 (B) has been prepared in a wide diversity of nanostructures, [3] or in hybridization with graphene materials to improve electronic conduction. [4] In general, nanostructured electrode materials are advantageous in achieving better kinetics of ion diffusion and charge transfer compared to bulk counterparts. There exist controversial mechanisms of the capacitive storage in TiO 2 (B) which is mainly due to the large variety of morphologies from different synthetic methods. [5] It is common that the synthesis procedures have a big impact on the morphology and subsequently the electrochemical performance. [6] On a microscopic level, the anisotropy of lithium-ion diffusion in crystal lattice also plays a significant role in the lithium-ion insertion and transportation. [7] It has been predicted that the lithium-ion coordination locates at the A2 site and C site of (001) surface, A1 site of the (010) and (100) substrate of TiO 2 (B). [8] Subsequently, the C site is further identified as C′ site due to the theoretical energy favorable tendency. [2a] Meanwhile, the different site sequences of the Li-ion coordination and lithiation process were proposed controversially by different techniques. For example, the sequence of C-A1-A2 is confirmed by neutron diffraction [9] and X-ray diffraction (XRD), [10] A2-C-A1 for bulk and C-A2-A1 for nanosheet are revealed by density function theory (DFT) calculation, [5b] and A2-C′-A1 is determined by the combination of in situ/operando X-ray pair distribution function (PDF) and galvanostatic intermittent titration (GITT) technique. [2b] Most of the previous calculations unraveled the coordination and thermodynamic insertion of lithium ion to TiO 2 (B), whereas few work has been done to elucidate the lithium-ion transportation in TiO 2 (B), especially the anisotropy of transportation, which would unravel the kinetics of lithiation and delithiation process of the electrodes.
In this work, we conduct first-principle calculations of the formation energy of Li x Ti 8 O 16 with the coordination (A2, C′, C, and

Lithium Coordination in TiO 2 (B)
The formation energy of intermediate composition has been calculated to predict the stable crystal structure in the lithium/ sodium-ion intercalation/deintercalation process, which is necessary for a better understanding and elucidation of the mechanism of lithium/sodium ion storage in the electrochemical reaction. [14] Herein, the possible lithium-ion coordination in TiO 2 (B) includes C, C′, A2 and A1 sites, was screened for the stable structure at the varied composition. As shown in Figure 1a, the lithium ion energetically prefers to occupy the C′ site rather than the C site at the concentration of 0.75, 1, and 1.25; and the A2 site instead of the A1 site at all the concentrations. Moreover, the stable lithiation compounds of Li 0.5 TiO 2 (A2), Li 0.75 TiO 2 (A2+C′), and the final Li 1. 25 TiO 2 (A2+C′+A1) were obtained with the lowest formation energy at the corresponding lithium-ion concentration, as shown in Figure 1b. The full lithiation of TiO 2 (B) with a theoretical capacity of 419 mAh g −1 agrees well with the first discharge capacity of TiO 2 (B) at an elevated temperature of 50 °C. [2b] In order to investigate the electrochemical behavior of lithiation in TiO 2 (B), the corresponding potential for the lithium-ion insertion was calculated on the possible four sites (Figure 1c). Interestingly, the lithium-ion insertion at the A2 site has the highest electrochemical potential of 1.21 V (vs Li + /Li) as compared with the other three sites, A1 site (1.08 V), C (1 V), and C′ site (1 V). This calculated profile supports our above proposal that the lithium ion would occupy the A2 site first, then combine with C′ site, and finally access the A1 site according to the thermodynamic formation energy calculation of Li x TiO 2 (x = 0-1.25). Moreover, the calculated Li-storage profile show that it can uptake 0.81 Li + per molecular with a plateau of 1.5 V (Figure 1c), which is close to the experiment value of 1 V at 0.1C. It indicates there is still a potential to improve the practical electrochemical capacity for a high energy density electrode materials.
The XRD pattern evolution during lithiation was also calculated (data presented in Figure 1d,e). The main changes occur in the range of 10°-30° and 40°-50°. The peaks in the range of 20°-25°, corresponding to the most intensive diffraction from (110) planes, vary both peak position and intensity with the increase in Li uptake. Two other changed XRD pattern sections located in the ranges of 25°-30° and 40°-50°, where the (002) and (401) peaks weaken gradually in intensity during the lithiation, and the (020) and (601) peaks strengthen vice versa. These variations in both 2θ position and peak intensity are in good agreement with the reported in situ XRD observation. [10] It is important to note the evident peak shift in the range of 12°-15°: the (001) and (200) peaks first shift closer and then apart, and the (001) peak becomes the strongest after full lithiation (x = 1.25).
The expansion of the crystal lattices a, b, and c is anisotropic. Both a and b axes increase by 4% at the lithium-ion concentration of x = 0.75, while the c decreases by about 2% for full lithiation, as shown in Figure S1 in the Supporting Information. This agrees well with the in situ XRD observation of the discharge process to Li 0.75 TiO 2 (B). [10] Eventually at the full uptake of Li (x = 1.25), b expands up to 11%, while a gets close to the original value a 0 . The volume expansion in TiO 2 (B) is less than 11% after the full lithium-ion intercalation, which assures the good cycle durability for lithium-ion batteries.

Lithium-Ion Transportation in TiO 2 (B)
To further understand the initial lithium-ion intercalation, the energy barriers of lithium-ion diffusion in Li 16 Ti 32 O 64 (A2) and Li 24 Ti 32 O 64 (A2+C′) were calculated and shown in Figure S2 and Table S1 in the Supporting Information. According to the symmetry of the crystal lattice, lithium ions first occupy the A2 site at high electrochemical potential, hopping within the 2D ab plane with a zigzag pathway along the b-axis (yellow arrows in Figure 2b) and zigzag-linear along a-axis (blacks arrows in Figure 2b). Kinetically, the zigzag hopping is rate-limiting for diffusion along with both a and b directions (Figure 2b), which gives a higher energy barrier of 0.88 eV in comparison with 0.44 eV in the linear hoping step (Table S1, Supporting Information). The lithium-ion diffusion coefficients of the zigzag and linear steps were calculated as 1.75 × 10 −17 and 5.48 × 10 −10 cm 2 s −1 at room temperature. The former limited the whole ions transportation process in the ab plane, which indicates a much sluggish kinetics at the initial lithiation stage (A2 site occupation).
As the lithium ion further occupies the central C′ site in the ab plane, the repulsion of lithium ion between central C′ and A2 sites (Figure 2c) increases the distance between two A2 sites, and leads to a high hopping energy barrier of 1.72 eV (Table S1, Supporting Information). Consequently, the linear hopping step switches to the zigzag pathway via the central C′ site which lowers the diffusion energy barrier to 0.6 eV (Figure 2d). This is comparable to 0.56 eV of the zigzag pathway along the b-axis. Accordingly, the lithium-ion diffusion coefficients for the two paths are also quite similar: 8.81 × 10 −13 and 3.0 × 10 −12 cm 2 s −1 at room temperature  (Table S1, Supporting Information). The lithium-ion diffusion coefficient is significantly enhanced at the lithiation stage of Li 0.75 TiO 2 , which agrees well with the result by the electrochemical GITT examination. [2b] The final lithium-ion insertion into the A1 site undergoes a hopping from A2 to A1 site on the {010} facet, or from C′ to A1 site on the substrate of the {100} facet. Insertion into the A1 site has been identified as retardation of lithium-ion diffusion in the electrochemical examination.
[2b] Therefore, we may conclude that, lithium ions prefer to intercalate through the {100} facet of TiO 2 (B) with relatively fast lithium-ion transportation rather than being adsorbed on {010} or {001} facets.

Synthesis and Characterization of TiO 2 (B)
The chemical reactions during the materials synthesis are depicted by Equations (S1)-(S3) in the Supporting Information.
The involved three chemical reactions were calculated by the enthalpy change in the value of −0.93, −2.16, and −1.75 eV. The calculation indicates that synthesis of single crystal of such materials is feasible (Table S2, Supporting Information). The belt morphology of H 2 Ti 3 O 7 , TiO 2 (B), and TiO 2 (B)@PP can be seen clearly from the scanning electron microscopy (SEM) images ( Figure S3a-c, Supporting Information) display the ultralong nanobelts. Interestingly, a bundle of TiO 2 (B) nanobelts that coincidently align in one orientation exhibits a single electron diffraction pattern, which shows that these nanobelts are single-crystalline with their axis along the [010] direction ( Figure S3d  the nanobelts are shown from which the low index (100) facet for the width, (010) facet at the termination and (001) facet for the thickness can be easily identified (Figure 3). Notably, to achieve TiO 2 (B) with a large exposure of (100) facets has been considered as a challenge due to the highest surface energy among the low index facets. [11b,c,15] In our study, based on the statistical measurement, the area of (100) facets in H 2 Ti 3 O 7 and TiO 2 (B) is estimated as 73.56% and 73.34% (width/thick ratio: 4/1), respectively. Notably, it is the first time to report such large exposure of {100} facets in these two materials. The large percentage of {100} facets in our TO 2 (B) may stem from a topotactic transformation from the H 2 Ti 3 O 7 precursor during the dehydrogenation. Similar topotactic transition from hydrogen titanate nanowire to TO 2 (B) has also been identified previously, where in situ HRTEM observation show that the crystal lattice underwent very small shrinkage in [100] axis direction and expansion in the [010] axis direction during the annealing process. [16] Furthermore, the nitrogen adsorptiondesorption isotherm was performed to obtain the specific surface area, which was calculated to be 48.77 m² g −1 using the Brunauer-Emmett-Teller (BET) method. The pore-size distribution (PSD) curve was also calculated ( Figure S4, Supporting Information), which implies that these TiO 2 (B) nanobelts may contain micropores (<2 nm).
The composition and phase of the TiO 2 (B)@PP composite are further characterized (Figure 4). The XRD spectra prove the monoclinic TiO 2 (B) phase (JCPDS: 46-1237). An intensive (110) Bragg reflection at 2θ of 24.9° is observed (Figure 4a). This peak is undistinguishable from the (101) diffraction from anatase TiO 2 at 25.3° if the latter is present. Note that TiO 2 directly grown by the common solvothermal method will contain more or less anatase TiO 2 . In this work, our two-step growth method (hydrothermal reaction followed by hydrochloric acid treatment) produces H 2 Ti 3 O 7 with high purity of monoclinic phase, which is consequently calcined into TiO 2 (B). The thermogravimetric analysis (TGA) test confirms that poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT-PSS) (PP) polymers in the composites with a weight ratio of 5.69% (Figure 4b). Fourier transform infrared spectroscopy (FT-IR) spectra of both the as-prepared TiO 2 (B) and TiO 2 (B)@PP display prominent bands at 3417 and 1610 cm −1 , which could be ascribed to O-H stretching. [17] The bands at 979 and 781 cm −1 correspond to the TiO stretching mode in crystalline TiO 2 (B) (Figure 4c). [18] The small features in the range from 1600 to 1000 cm −1 are the characteristic bands of PP. In the enlarged spectra ( Figure 4d) the high-intensity bands at 1517, 1328, 1147, and 1089 cm −1 reflect the CC, CC, SO, and S-phenyl vibrations, respectively. The low-intensity features located at 1205 and 1132 cm −1 are the stretching vibration of COC. [19]  X-ray photoelectron spectroscopy (XPS) characterization is useful to analyze the covalence state of PP on the TiO 2 (B) surface. The XPS survey spectrum reveals that the product is composed of Ti, O, C, and S elements (Figure 4e). The Ti 2p 1/2 and 2p 3/2 have distinctive binding energies at 464.1 and 458.4 eV, respectively, which correspond to the Ti 4+ ions (Figure 4f). [20] The O 1s spectrum is dominated by the TiOTi bond at 529.7 eV in TiO 2 (B); the two shoulders due to COC bond at 529.4 eV and SO bond at 531.5 eV are from the SO 3 2− function group of PP (Figure 4g). [21] The C 1s spectrum can be deconvoluted to three peaks: the main peak at 284.7 eV is for CC/CC group, and the two minor peaks are for COC (284.9 eV) and CSC (285.4 eV) groups (Figure 4h). [22] The S species are from the PSS (170-165 eV) and PEDOT (165-160 eV) (Figure 4i). [22,23]

Electrochemical Performance of TiO 2 (B)@PP
The electrochemical performance of the single-crystalline TiO 2 (B)@PP nanobelts and pristine TiO 2 (B) was evaluated, as shown in Figure 5. To identify the protective effect of PP, all the electrochemical measurements were performed with different voltage cut-off interval of 1-3, 0.5-3, and 0.05-3 V. The galvanostatic discharge/charge profiles at 0.1C (current: 33.5 mA) for TiO 2 (B)@PP are shown in Figure 5a. With the conductive and protective PP coating, the TiO 2 (B)@PP can be discharged down to 0.5 and 0.05 V, where the capacity is enhanced significantly from the 270.4 mAh g −1 (1-3 V) to 340.8 mAh g −1 (0.05-3 V) in the first discharge/charge cycle. The TiO 2 (B)@PP electrodes can reach the theoretical capacity of LiTiO 2 (≈335 mAh g −1 ) at room temperature. Moreover, for the first cycle of TiO 2 (B)@PP, Figure 5a shows that the charge and discharge capacities nearly coincide in all tested voltage ranges. This means the Coulombic efficiency approaches 100%. In contrast, the pristine TiO 2 (B) electrodes have a much lower first-cycle Coulombic efficiency of 97.6%, 84.7%, and 71.6% with cut-off potential of 1.00, 0.50, and 0.05 V, respectively ( Figure S5a, Supporting Information).
To identify the stability of cycled voltage profiles, the 1st, 300th, and 600th discharge curves of TiO 2 (B)@PP hybrid electrode at 1C are presented in Figure 5b. In the potential range of 1-3 and 0.5-3 V, the voltage profiles of discharge keep their initial shape with a steady voltage plateau at 1.5 V. The capacities at the 600th cycle are 179.6 mAh g −1 (1-3 V) and 214.8 mAh g −1 (0.5-3 V), corresponding to a capacity retention of 90% and 87%, respectively. As for the discharge potential range of 0.05-3 V, a higher capacity of 265.6 mAh g −1 is retained at the 600th cycle. However, the sloping of the discharge curve indicates enhanced intercalation pseudocapacitive behaviors. As for the pristine TiO 2 (B) electrode, it suffers sharp fading in capacity during cyclic discharging to 0.5 and 0.05 V ( Figure S5b, Supporting Information). This indicates that the PP coating has a protective effect to TiO 2 (B).
The protective effect of PP can also be manifested from the improved rater capability of the TiO 2 (B)@PP electrode (see Figure 5c and Figure S5c in the Supporting Information). With the protective PP, the relatively high gravimetric capacity for the low discharge voltage cut-off at 0.5 and 0.05 V are consistent at all different rates from 0.1C to 30C. It indicates the TiO 2 (B)@PP can achieve high capacities when discharged to 0.5 V as there is no redox reaction between Ti 4+ species and electrolyte. [24] For the pristine TiO 2 (B), however, during the initial five cycles at the rate of 0.1C, the capacity fades dramatically when discharged to 0.5 and 0.05 V, even to the high voltage cut-off of 1 V ( Figure S5c, Supporting Information). A comprehensive summary of the capacity at different cut-off potentials and current rates is illustrated in Figure 5d, from which we can clearly see the enhancement of capacity due to PP protection. The rate capability of our single-crystal TiO 2 (B) exceeds those from other reports (see Figure S6 in the Supporting Information). This could be ascribed to the high chemical activity {100} surfaces of our TiO 2 (B) nanobelts for lithium-ion insertion and diffusion. Additionally, the conductive and protective PP coating enhances the transfer kinetics and prevents direct contact with the electrolyte.
The protection effect of PP can be further confirmed by the long cycle life test (Figure 6). The TiO 2 (B)@PP electrode delivers a capacity of 245.7 mAh g −1 when the electrode is discharged to 0.5 V (Figure 6a), while the capacity decreases to 209 and 240.3 mAh g −1 at the discharging potential of 1 and 0.05 V, respectively (Figure 6b). After 1500 cycles of charge and discharge in the range of 0.5-3 V, the capacity of TiO 2 (B)@PP lowers to 160.7 mAh g −1 , with a retention ratio of 66%. For comparison, for the potential rage of 3-1 V, the electrode can maintain a stable capacity, but the value is low. Notably, the capacity of TiO 2 (B)@PP with discharging potential down to 0.05 V undergoes a slight decay to about 200 mAh g −1 at the 200th cycle, and then gradually increases to 310 mAh g −1 after 900 cycles (Figure 6b). This is probably caused by the reduction of Ti 4+ to Ti 0 and the formation of Li 2 O. [25] As for the pristine TiO 2 (B) electrode, a stable capacity can be maintained when it is discharged to 1 V, but it sharply fades in the case of discharging potential of 3-0.5 V ( Figure S5d, Supporting Information), manifesting its less stability than the TiO 2 (B)@PP electrode.

Lithium Diffusion Kinetics
The electrochemistry of the TiO 2 (B)@PP electrode was further examined by the cyclic voltammetry (CV) with different scan rates in order to understand the kinetic behavior (Figure 7a). The main peaks with two small branches at both oxidation and reduction sides can be observed when the scan rate is small. With increasing the scan rate, the cathodic and anodic peaks shift and broaden due to polarization effect. The power-law relationship between the peak currents (i) and the scan rate (v) is plotted in logarithm scale, which shows a characteristic linear proportionality. [26] The gradient of the plots are 0.41 and 0.35 for cathodic and anodic peaks at the small scan rate of ≤0.5 mV s −1 , whereas it increases to 0.79 and 0.75 at the scan rate of ≥0.6 mV s −1 (Figure 7b). This confirms our proposal that the lithium ion storage in TiO 2 (B) is dominated by the diffusion-limited process. The increase in gradient at higher scan rates is due to polarization effect, namely, the ion transportation cannot catch up the potential sweep due to limited kinetics of the lattice diffusion.
Finally, to evaluate the lithium-ion diffusion coefficient at the initial stage of the discharging process, the electrochemical impedance spectrum (EIS) examination of pristine TiO 2 (B) without polymer coating was performed after three discharge/ charge cycles at 0.1C, as shown in Figure 7c. The EIS simulation was carried out by Zview software, and the Warburg factor fitting was also obtained (Figure 7d). Both EIS simulation and Warburg factor fitting agree well with the experiment results. The initial lithiation stage (less than 0. 5 Li) in TiO 2 (B) has a sluggish kinetics and low diffusion coefficient according to our first-principle calculations. The lithium-ion diffusion coefficient is calculated to be 1.75 × 10 −17 cm 2 s −1 , approaching 1.82 × 10 −15 cm 2 s −1 obtained from simulation of EIS spectrum (Table 1) and 1.79 × 10 −16 cm 2 s −1 determined by GITT calculation ( Figure S7, Supporting Information). At the lithiation stage of the A2+C′ site, the lithium-ion diffusion coefficient rises to 8.81 × 10 −13 cm 2 s −1 (see Table S1 in the Supporting Information). This implies that, when the kinetics of intercalation improves, the lithium-ion diffusion plays the dominating role.

Conclusion
We have established that TiO 2 (B) with large (001) facets exposure can store 1 lithium ion per molecular with specific capacity up to 335 mAh g −1 , and even possibly up to the theoretical capacity of 419 mAh g −1 (1.25Li + ). This is in contrast to the anatase TiO 2 anode that has been widely studied for lithium ion storage. First-principle calculation proposes a 2D diffusion pathway for lithium-ion insertion and transportation through {100} facets of TiO 2 (B). Experimentally, we have achieved the synthesis of TiO 2 (B) nanobelts with unusual {100} facts exposure. After coating with a thin layer of copolymer, the obtained TiO 2 (B)@PP nanobelts electrode exhibits enhanced electrochemical kinetics with high capacity and long cycle-life for lithium ion storage. This work provides a new strategy to screen and design nanoarchitecture with ideal orientation   based on the anisotropy of lithium-ion diffusion in the crystal lattice. It could be extended to many other metal oxides that have a large variety of crystal phases for battery applications, including but not limited to manganese oxide, vanadium oxide, and iron oxides.

Computational Methods
All calculations were based on DFT ultrasoft pseudopotential methods within the CASTEP module [27] in Materials Studio. The generalized gradient approximation (GGA) of Perdew-Becke-Ernzerhof (PBE) [28] with ultrasoft pseudopotential [29] was used for the exchange-correlation function. The Brillouin zone was sampled by a 1 × 4 × 2 Monkhorst-Pack k-points grid. The optimized structures were obtained by relaxing all the atomic configurations until interatomic forces were less than 0.01 eV Å −1 and energy change of convergence were less than 10 −5 eV per atom, with a kinetic-energy cutoff of 400 eV.
The lithium-ion diffusion calculations were performed on the DMol 3 code. [30] The generalized gradient-corrected Perdew-Burke-Ernzerhof functional [28] applied for all calculations, along with a double numerical basis set. The double numerical basis set considers a polarization d function on heavy atoms and a polarization p function on hydrogen atoms. During the coordinates relaxation, the tolerances of energy and force set values of 1 × 10 −5 Ha and 0.002 Ha Å −1 , respectively, and the maximum displacement placed at 5 × 10 −3 Å. The Monkhorst-Pack k-point mesh was 1 × 2 × 1 for the supercell of 1 × 2 × 2.

Structure and Theory
The voltage for lithium-ion insertion was calculated in the sequence of lithium-ion coordination site by site. Specifically, the voltage, relating to the Li chemical potential according to the Nernst equation, could be expressed as follows where μ cathode , the chemical potential of the intercalation compound, is the derivative of the free energy for Li concentration x, μ anode is the chemical potential of the reference anode (a constant for metallic Li), and F is the Faraday constant. The free energy change (ΔG) can be calculated approximately by the internal energy change per intercalated Li + , E(x) is the total energy of Li x TiO 2 , which is an average value for all Li x TiO 2 compositions between x 1 and x 2 (x 2 > x 1 ). E (Li) represents the total energy of lithium atom.
To investigate the lithium-ion intercalation and diffusion in TiO 2 (B), single vacancy migration in a supercell of [Li 0.5 TiO 2 ] 32 (A2) , and [Li 0.75 TiO 2 ] 32 (A2+C′) was calculated by the transition state (TS) module in Dmol 3 code. The TS of the single vacancy migration, located on the hypersurfaces of the potential energy, is calculated by performing a linear synchronous transit (LST), quadratic synchronous transit (QST), and conjugate gradient refinements. [31] An LST optimization calculation was carried out first, followed by QST maximization with conjugate gradient minimization to obtain the transition state. The cycles repeated to reach a stationary point, which is recognized as the transition state.

Experimental Section
Materials Synthesis: The synthesis method of single-crystal TiO 2 (B) was modified from the typical hydrothermal method to obtain hydrogen titanate, and subsequently, the synthesized hydrogen titanate was calcined. Generally, the sodium hydroxide (NaOH, AR, 96%) solution (10 mol L −1 , 300 mL) and 2 g titanium dioxides (TiO 2 , Anatase, CP, 98%) were mixed with vibration at a given speed by mechanical equipment at 165 °C for 48 h. Then, the mixture was dropped by hydrochloric acid (HCl, AR, 37%) until the pH value reached to 1. The precipitate of single-crystalline hydrogen titanate (H 2 Ti 3 O 7 ) was generated. Finally, the single-crystal TiO 2 (B) was obtained by calcining the hydrogen titanate at 400 °C for 4 h.
Materials Characterization: The morphology and crystalline structure of the as-prepared single-crystal H 2 Ti 3 O 7 and TiO 2 (B) were characterized by scanning electron microscope (Hitachi-S4800, Tokyo, Japan), high-resolution transmission electron microscope (FEI Tecnai G2F30, Hillsboro, Oregon, USA), and X-ray diffractometer (XRD-6100, Kyoto, Japan). Meanwhile, FT-IR (Nicolet-iS5, Thermo Fisher Scientific) was used to characterize the presence of PEDOT:PSS. The shift of binding energy was determined by XPS ( Escalab 205Xi, Thermo Fisher Scientific). The weight ratio of coated PEDOT-PSS was measured by thermal gravimetric analyzer (JOEL-JEM-2100, Tokyo, Japan). In addition, nitrogen adsorption and desorption isotherms were performed on a Micromeritics ASAP 2460 instrument at 77 K. The value of specific surface area was calculated by the BET method. The PSDcurve was determined using the Barrett-Joyner-Halenda (BJH) algorithm.
Electrochemical Examination: Electrochemical properties were measured by fabricating CR2032-typed coin cells with lithium metal counter electrode in Ar-filled glove box. The electrode films were prepared with a composition of 80 wt% active material, 10 wt% acetylene black, and 10 wt% polyvinylidene fluoride (PVDF) binder. The electrolyte was a 1 mol L −1 LiPF 6 in ethylene carbonate (EC), dimethyl carbonate (DMC), and diethyl carbonate (DEC) solution, EC/DMC/DEC = 1:1:1 vol%. The galvanostatic discharge/charge tests were carried out using a Land battery test system CT-2001A (Wuhan, China) in the voltage ranges of 0.05-3, 0.5-3, and 1-3 V, respectively, versus Li counter electrode. The GITT examination was also performed on the Land battery test system CT-2001A (Wuhan, China), discharging at the current of 0.1C. CV curves were recorded from 0.5 to 3 V at scan rates of 0.1, 0.3, 0.5, 0.6, 0.7, 0.8, and 0.9 mV s −1 . The EIS was carried out within the frequency range of 0.01 Hz to 100 kHz after three discharge/charge cycles at 0.1C. Both CV and EIS were tested by CHI600D electrochemical workstation (Shanghai, China).

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.