Electrochimica Acta 182 (2015) 368–375
Contents lists available at ScienceDirect
Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta
Gd doped single-crystalline Li4Ti5O12/TiO2 nanosheets composites as superior anode material in lithium ion batteries Yueming Li* , Zhiguang Wang, Dan Zhao, Long Zhang State Key Laboratory of Metastable Materials Science and Technology, College of Materials Science and Engineering, Yanshan University, Qinhuangdao, Hebei Province, 066004, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 18 August 2015 Received in revised form 17 September 2015 Accepted 18 September 2015 Available online 25 September 2015
Spinel Li4Ti5O12 has been thought as one of ideal anode materials for Li ion batteries due to its safety performance and long cycle life, while the low conductivity and Li-ion diffusion capability has limited its application. To improve the electrochemical performance of Li4Ti5O12, Gd-doped single-crystalline Li4Ti5O12/TiO2 (rutile phase) nanosheets composites in this study was prepared via a solvothermal synthesis assisted method. It is found that the as prepared doped dual phases composites showed superior electrochemical performance as anode material. At a current density of 0.2 A/g (1.14 C), the GdLTO/TiO2 electrode can deliver a speciﬁc capacity of 180.2 mAh/g, and it can deliver a speciﬁc capacity as high as 111.1 mAh/g even at current density of 20 A/g (114.3 C). The introduction of Gd ion has been found to play a crucial role in improving the electrochemical performance, and its introduction modiﬁes the morphology of as prepared sample at the same time. ã 2015 Elsevier Ltd. All rights reserved.
Keywords: Lithium ion batteries Doping Li4Ti5O12 rutile nanosheets
1. Introduction Rechargeable Li ion batteries (LIB) has become one kind of dominant battery which powers most of today’s portable electronics due to their high energy density, environment friendly nature, as well as the good cycle life [1,2]. However, it is a key issue to develop safer LIB with quick charge/discharge capability especially for high power appliance like electric vehicles (EV). As is known, the commercially used anode material in most Li-ion batteries is graphite, which may bring the safe issue due to the decomposition of the organic electrolyte at the low discharge potential . In addition, most of anodes materials with high theoretical capacity in LIB like Si and SnO2, still suffer from insufﬁcient rate performance and short cycle life due to the serious volume expansion during charge/discharge process . It is necessary to develop the high performance anode materials with high safety. Spinel Li4Ti5O12 has been thought as one of ideal anode materials in LIB with a high safety performance due to its high charge-discharge plateau potential (about 1.55 V versus lithium).
* Corresponding author. Tel.: + 86-335-8064637. E-mail address: [email protected]
(Y. Li). http://dx.doi.org/10.1016/j.electacta.2015.09.103 0013-4686/ ã 2015 Elsevier Ltd. All rights reserved.
The high plateau potential helps to alleviate the formation of the solid electrolyte interface (SEI), thus a typical issue in carbon based anodes can be avoided . Furthermore, spinel Li4Ti5O12 belongs to the zero-strain insertion material, and the volume change is almost negligible during the insertion/extraction for Li ions . Although spinel Li4Ti5O12 as an electrode material in LIB has its own advantages, its poor electronic conductivity (about 10 13 Scm 1) and low Li-ion diffusion ability would limit its quick charge-discharge ability as anode materials, which is of great importance for the application in high power appliances like EV. Up to now various efforts have been made to improve the electronic conductivity and/or Li-ion diffusion ability of spinel Li4Ti5O12. In general, strategies including synthesizing materials with smaller particle size [6–14], element doping [8,15–22] and surface modiﬁcation with highly conductive additives,[23–28] have been proposed. The conductive additives to improve the conductivity of LTO electrode can be divided into two main categories, that is, carbon-based and metal-based materials (metal or metal oxide nanoparticles). Carbon coating has been proved as a very effective method to improve the electrochemical performance of LTO as well as other metal oxide due to the high conductivity of carbonaceous material based on our previous work and others’ work [29–32]. However, the introduction of carbon will reduce the tap density of LTO, thus lowering the practical energy density for
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
the cell using LTO electrode. Furthermore, the introduction of carbon materials will also lead to the irreversible capacity during the ﬁrst discharge due to the formation of SEI. Metal nanoparticles such as Ag nanoparticles are very good candidates to improve the conductivity of LTO . However, these metal nanoparticles such as Ag nanoparticles cannot contribute speciﬁc capacity within the potential window of LTO, leading to decreased speciﬁc capacity and increased preparation costs. Furthermore, the multiple steps in general were needed to coat metal particles on LTO, leading to the rising costs in preparation. Certain metal oxides have their own advantages as coating material for LTO because they can not only improve the ionic or electronic conductivity of LTO, but also store Li ions which will contribute the speciﬁc capacity to the whole batteries . Furthermore, the introduction of metal oxide will not lower the density of LTO based composites due to the high density for metal oxide, ensuring the high volumetric energy density for LTO/metal oxide composites. For example, Wang et al. has reported Li4Ti5O12 nanosheets terminated with rutile-TiO2 at the edges exhibited much better electrochemical performance than pure LTO in Li-ion batteries . For the strategy of reducing the particle size, the advantage using Li4Ti5O12 with smaller particle size (including nanostructured Li4Ti5O12) is that the Li-ion insertion/extraction pathway can be shortened, leading to greatly improved rate performance . Notably, the conductive additives coating affects the electron transport only on the surface of active electrode materials, while the intrinsic ionic and electronic conductivities of LTO have not been enhanced. It has been proved that element doping to LTO is an effective method to improve its intrinsic ionic and electronic conductivity based on the experiments and theoretical simulation using density functional theory . The combination of heteroatom doping, metal oxide coating and decreasing the particle size for LTO would further improve its electrochemical performance based on above analysis. However, to the best of our knowledge, there are little such reports on nanostructured LTO with both element doping and metal oxide coating as electrode in Li-ion batteries. In this study, we have synthesized Gd-doped LTO nanosheets with rutile TiO2 coating, and the electrochemical experiments showed as prepared products exhibited superior performance as anode materials in Li-ion batteries. 2. Experimental 2.1. Materials All reagents were of analytical-grade reagents and used without further puriﬁcation except otherwise stated. Commercial LTO was purchased from Hefei Kejing incorporation. 2.2. Synthesis of Gd-doped LTO/TiO2 composites Gd-doped LTO/TiO2 nanocomposites were prepared according to previous reports with a little modiﬁcation.[6,35] In a typical synthesis procedure, 2.0 ml of tetrabutyl titanate, 0.03 g of GdCl36H2O, and 0.19 g of LiOHH2O were thoroughly mixed in 20 ml of ethanol at room temperature. Then, 25 ml of deionized water were added under stirring. After stirring for 4 h, the mixture was transferred to Teﬂon-lined stainless autoclave and placed in an oven at 180 C for 48 h. The white powder was ﬁltered, washed with deionized water until the ﬁltration become neutral and dried at 80 C overnight. Finally, the white product was heated at 700 C for 6 h in a horizontal tube furnace in air to obtain the Gd doped LTO/TiO2. For comparison, undoped LTO/TiO2 was prepared similar to the above procedure except no GdCl36H2O was added.
2.3. Characterizations The powder X-ray diffraction (XRD) measurements of the samples were recorded on a D-max 2500 X-ray powder diffractometer using a graphite monochromator with Cu Ka radiation (l=1.5406 Å). The data were collected between scattering angles (2u) of 10–80 at a scanning rate of 4 /min. Scanning electron microscopy (SEM) was carried out on Hitachi S-4800. Transition electron microscopy (TEM), high resolution TEM and selected area electron diffraction (SAED) were performed on JEM 2010 transition electronic microscopy and FEI Tecnai F20 G2 STWIN. TEM specimens were prepared by drop-casting the as prepared sample dispersions onto carbon-coated TEM grids and dried in air. The Brunauer-Emmett-Teller (BET) speciﬁc surface area was calculated from N2 adsorption/desorption isotherms which were obtained by a gas adsorption analyzer (ASAP 2020, Micromeritics Instrument Co. USA) at 77K. X-ray photoelectron spectroscopy (XPS) was carried out on ESCALAB 250XI and the binding energy is calibrated with C1s=284.8 eV. 2.4. Electrochemical tests The electrodes were constructed by mixing the active materials (80 wt%), conductive carbon black (15 wt%) and polyvinylidene ﬂuoride (PVDF) binder (5 wt%) homogeneously in N-methyl pyrrolidinone (NMP) solvent and spread onto copper foil. The electrode was dried under vacuum at 120 C for 12 h before electrochemical evaluation. The loading density for the active materials in the electrode is about 1 mg/cm2. The electrolyte employed was 1 M LiPF6 in an ethylene carbonate (EC) and dimethyl-carbonate (DMC) solvent (1:1, v:v). The cells were assembled inside the argon-ﬁlled glove box (Braun, H2O <0.5 ppm and O2 < 0.5 ppm) using a lithium metal foil as the counter electrode and the reference electrode and microporous polypropylene as the separator. The galvanostatic charge and discharge of the assembled cells were tested on a BT 2000 battery testing unit (Arbin Inc., USA) at varied current densities between the voltage limits of 1.0 and 3.0 V (vs. Li/Li+). 1 C is deﬁned to the current required to charge/discharge the battery in one hour, and 175 mA/g for LTO. The cyclic voltammetry (CV) measurements were carried out in the potential range of 1.0 to 3.0 V (vs. Li/Li+) on an electrochemical workstation (P4000, Princeton Applied Research, USA) at varied scanning rates. Electrochemical impedance spectroscopy (EIS) was carried out using a P4000 electrochemical workstation within a frequency range of 0.1–105 Hz at the amplitude of 5 mV versus the open circuit potential. 3. Results and Discussion We demonstrate here the preparation of Gd doped LTO/TiO2 and their superior electrochemical performance in Li-ion batteries. We rely on Gd doping and TiO2 coating to modify the electrochemical performance of LTO, in this way that Gd ions served as heteroatoms and facilitated the formation of nanosheets, while TiO2 helps to improve the diffusion coefﬁcient.. Fig. 1 shows the XRD patterns of as prepared samples and commercial LTO. For the commercial LTO, the sharp and strong peaks located at 18.4, 35.6, 43.3, 48.2, 57.2, 62.7 and 66.3 can be indexed to the (111), (311), (400), (331), (511), (440) and (531) diffractions of cubic spinel Li4Ti5O12 (JCPDS card 49-207). The XRD patterns for undoped LTO/TiO2 and Gd-doped LTO/TiO2 composites not only showed typical characteristics peaks of LTO but also showed the characteristics peaks of tetragonal rutile TiO2 (JCPDS card no. 21-1276), indicating the presence of LTO as well as rutile TiO2. The characteristic peaks of both LTO and TiO2 in the composites are sharp and strong, suggesting the high crystallinity
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
using the BET equation was 9.4 m2g 1, while that of Gd-doped composites increased to 14.9 m2g 1, consistent to the thinner morphology. TEM was used to further clarify the structure of the Gd-doped composites. It is clearly visible that Li4Ti5O12 was composed of nanosheets with the irregular shapes, consistent to the SEM observation. The TEM results indicated that thickness of the partial nanosheets is less than 10 nm. As shown in Fig. 2E, the spaces of the perpendicular lattices for the Gd doped sample derived from the HRTEM image are 0.48 and 0.30 nm respectively, which are in
Fig. 1. XRD patterns of commercial LTO, undoped LTO/TiO2 and Gd-doped LTO/TiO2 (triangle: rutile TiO2; square: LTO).
for these phases. The fraction of LTO is estimated to be 70% in in weight using Maud software . Fig. 2(A) and (B) presents the morphologies and particle sizes of commercial LTO and Gd doped LTO/TiO2 composites. The SEM image of commercial LTO clearly showed the irregular bulk particles with diameter in the range of hundred nanometer to micrometers. As shown in Fig. S1, the undoped LTO/TiO2 was composed of plenty of nanosheets with irregular shapes. The thickness of as formed stacked structure is about dozens of nanometers. The morphology of Gd-doped sample is similar to that of undoped one, while it seems that the particle thickness for Gddoped samples is thinner compared with that of undoped one, which is consistent to the previous reports on Gd-doped LTO . The speciﬁc surface area (SSA) of undoped LTO/TiO2 evaluated
good agreement with the d-spacing of (111) and (202) planes of spinel Li4Ti5O12. The selected-area fast Fourier transform (FFT) on the HRTEM image results in the FFT pattern of the cubic LTO structure (inset of Fig. 2E), which suggests the single-crystalline nature of LTO nanosheets. The selected-area electron diffraction (SAED) pattern (Fig. S2) shows the typical diffraction spots corresponding to the cubic phase consistent to XRD results, which further proved the single crystalline nature with a high crystallinity. SAED was also used to study the crystallinity of TiO2 in the composites. As shown in Fig. 3(C), the typical diffraction spots corresponding to the tetragonal phase of TiO2 were observed, indicating the single crystalline nature of rutile. The HRTEM in Fig. 3(D) showed that the spaces of the two lattices (plane angle of 38.8 ) are 0.33 and 0.17 nm, respectively, which are in good agreement with the d-spacing of (110) and (121) planes of rutile TiO2. The high crystallinity of rutile TiO2 was also conﬁrmed from the distinct spot patterns with the characteristics of tetragonal phase structure obtained by the FFT results, which is consistent to the SAED result. Fig. 4 shows the XPS spectrum of the Gd doped LTO/TiO2 composites. The survey XPS spectrum shown in Fig. 4(A) reveals the presence of Ti, and O in as prepared Gd-doped LTO/TiO2 composites. The presence of Li can be conﬁrmed by the high resolution spectrum of Li 1s, shown in Fig. S3. Figure 4 (B) illustrates the high-resolution XPS spectra of Ti 2p. Two prominent peaks located at about 458.7 and 464.6 eV can be
Fig. 2. SEM images of commercial LTO (A) and Gd-doped LTO/TiO2 (B); TEM images of Gd-doped LTO/TiO2 at low magniﬁcation (C) and at high magniﬁcation (D); HRTEM image of Gd-doped LTO/TiO2 (LTO crystal grain viewed from (121)) (E) (inset: FFT derived from HRTEM).
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
Fig. 3. TEM images of Gd-doped LTO/TiO2 at low magniﬁcation (A) and at high magniﬁcation (B); the corresponding SAED pattern of TiO2 crystalline grain in the sample (C) and HRTEM of TiO2 crystalline grain (D) (inset of D, FFT derived from HRTEM).
Fig. 4. The survey XPS of Gd-doped LTO/TiO2 (A); High resolution XPS spectra (Ti 2p) (B), O 1s (C) and Gd 4d(D)
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
assigned to Ti 2p3/2 and Ti 2p1/2 of Ti4+, respectively. The O 1s spectrum of Gd-doped LTO/TiO2 is shown in Fig. 4(C) with their peak curve-ﬁtting lines with respect to the chemical states. The O 1s peak at 530.1 eV represents the Ti-O bond in LTO and/or TiO2 , and the O 1s peak at 532.2 eV assigns to the chemisorbed oxygen on the surface of crystallites. Fig. 4(D) shows the Gd 4d high resolution spectrum. The broad and weak peak for the Gd 4d photoemission arises from multiplet splitting of the 4d hole with 4f7 valence electrons to form 9D and 7D ﬁnal ionic state . The doping amount of Gd element in as prepared LTO/TiO2 is 0.3 % (atomic ratio) based on the XPS experiment. Fig. 5 shows the cyclic voltammogram of the undoped LTO/TiO2, and as prepared Gd-doped LTO/TiO2 electrodes within a potential window of 1–3 V (vs Li/Li+). An obvious pair of peaks can be observed either in LTO/TiO2 or as prepared Gd-doped LTO/TiO2 electrodes at varied scan rate, which can be attributed to the redox reaction Ti4+/Ti3+ corresponding to the lithium insertion/extraction in the spinel lattice. It can be seen that the CV curves of Gd doped LTO/TiO2 have relatively stable shape with the increase of scan rate compared with un-doped electrodes and the commercial LTO electrodes (Fig. S4), indicating the better rate performance for Gddoped LTO/TiO2 electrode. The shapes of redox peaks observed in a CV plot can reveal the electrochemical reaction kinetics of Li+ insertion/extraction. In general, the sharp and well resolved redox peaks indicate the fast kinetics, while the broad redox peaks imply a slow process. As shown in Fig. 5(C), the peak shape of Gd doped LTO/TiO2 was sharpest among three electros, and the gap (0.132 V) between redox peaks was smaller than those of the un-doped (0.171 V) and the commercial LTO (0.232 V) electrodes at scan rate of 0.1 mV/s, indicating that the Gd doped LTO had faster kinetics and better efﬁciency of the redox reaction. Furthermore, the peak currents of the Gd doped LTO/TiO2 electrodes are highest among
these three electrodes, demonstrating much improved lithium diffusion and better lithium storage capability for Gd-doped LTO electrode. As shown in Fig. 5(D), it can be observed that the currents (Ip) of the cathodic peaks present a linear relationship to the square root of the scanning rate (n1/2), suggesting a diffusion-controlled reaction . Notably, the slopes of the cathodic peak for the Gd doped LTO/TiO2 are higher than those for commercial LTO and undoped LTO/TiO2, indicating the higher diffusion coefﬁcient for Gd-doped LTO electrode. For the diffusion controlled reaction, the peak current can be expressed to the following equation (T=298.15K). Ip=2.686105 n3/2 A D
Where Ip is peak current, n is charge transfer number, A is the surface area in the electrode, D represents the diffusion coefﬁcient, C is bulk concentration of the ions in the electrode and n represents the number of transfer electrons. According to Randles-Sevchik equation, the apparent diffusion coefﬁcient of Li ions for Gd doped LTO/TiO2 electrode is about 2.2910 10 cm2/s at 2 mv/s, which is much larger than those of undoped LTO/TiO2 (9.3610 11 cm2/s) and the commercial LTO (1.5110 11 cm2/s) at the same scan rate. These results demonstrate that the combination of Gd doping and Rutile TiO2 can effectively improve the electrochemical reaction kinetics of Li+ insertion/deinsertion. Fig. 6 shows the charge/discharge proﬁles of commercial LTO, undoped LTO/TiO2 and Gd doped LTO/TiO2 at varied current density. From these charge/discharge proﬁles shown in Fig. 6(A, B, and C), it can be observed all these electrodes showed a relative ﬂat charge/discharge plateaus at the potential around 1.55 V at low rates. Notably, the Gd-doped LTO/TiO2 electrode exhibited a relatively ﬂat potential plateau even at current density of 20 A/g
Fig. 5. Cyclic voltammograms of the undoped LTO/TiO2 (A), Gd-doped LTO/TiO2 (B) at various scanning rates from 0.1 to 5.0 mV/s; The comparison of CV plots at 0.1 mV/s (C), and the cathodic peak currents against square roots of scan rate (D) for the commercial LTO, un-doped LTO/TiO2 and Gd-doped LTO/TiO2 electrodes.
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
Fig. 6. Charge-discharge curves of the commercial LTO (A), the undoped LTO/TiO2 (B) and Gd-doped LTO/TiO2 (C) at varied current density; Rate performance (D) and cycle performance plots (E) of the commercial LTO, undoped LTO/TiO2, Gd-doped LTO/TiO2; Comparison of rate capability of Gd-doped LTO/TiO2 with some other LTO with high performance (F).
(114.3 C), while commercial LTO electrode exhibited a short and sloping curve at current density of 5 A/g (28.6 C). Although all the gap values between the charge and discharge plateau potential for these electrodes increased with increase of charge/discharge rates, but the values for each electrode varied widely. As shown in Fig. S5, the gap values for these three electrodes at 1 C rate are close, while they are signiﬁcant different at high rates. The Gd-doped LTO/TiO2 exhibited the least increase in gap values while the commercial LTO showed the largest increase among these electrodes at higher rates, indicating the low polarization and enhanced reversibility at higher rates for Gd-doped LTO/TiO2. The commercial LTO electrode can deliver a speciﬁc capacity of 149.5 mAh/g at the current density of 0.2 A/g, while undoped LTO/ TiO2 and Gd doped LTO/TiO2 can deliver a speciﬁc capacity of 159.9 and 180.2 mAh/g at the same rate, respectively. Considering the superior performance at low current density, higher rate performance was also studied. It is found that the discharge capacities for the commercial LTO electrode decrease rapidly with increase of current density. The commercial LTO electrode can only deliver a speciﬁc capacity of 25.2 mAh/g at 5 A/g (28.6 C). In contrast, the undoped LTO/TiO2 and the Gd-doped LTO/TiO2 electrodes showed much improved rate retention capability compared with commercial LTO electrode, while Gd doped electrode showed better performance than the undoped electrode. The Gd doped electrode can deliver a discharge capacity of 155 mAh g 1 at 2A/g (11.4 C) and 145.8 mAh/g at 5 A/g (28.6 C). Even at a high current density of 20 A/g (114.3 C), the composites electrode could deliver a speciﬁc capacity of 111.1 mAh/g, suggesting superior rate capability. As shown in Fig. 6(F), the speciﬁc capacity at lower rate is close to reported values. Notably, the discharge capacities at high rates (145.8, 136.1 and 114 mAh/g at 28.6, 57.1 and at 114.3 C,
respectively) are much better than previous reports on high performance LTO including Gd doped LTO (123.6 mAh/g at 10 C) 36, rutile coated LTO (110 mAh/g at 60 C)  C encapsulated F doped LTO (91 mAh/g at 60 C)  and Ag modiﬁed LTO nanosheets (140 mAh/g at 30 C).  It is reported that self-supported LTO is able to deliver a considerable capacity at higher rate (78 mA/g at 200 C) . However, there are several disadvantages for selfsupported LTO. The electrode of self-supported LTO needs careful control during the preparation process, and it is also hard to be prepared in a large scale. Furthermore, the very low loading (less than 0.1 mg/cm2) of LTO will lead to the low volumetric energy density, which is of critical importance in the practical applications. The cycle stability has been studied for 200 cycles at 0.2 A/g (1.14 C) and 500 cycles at 5 A/g (28.6 C), shown in Fig. 6(E). The capacitance can retain 96.1% of its initial speciﬁc capacity at 0.2 A/g, and the capacitance show negligible decay at the following 500 cycles at 5A/g (28.6 C), indicating the excellent cycle stability for Gd doped LTO/TiO2 electrode. Although there are rutile TiO2 existing in the composites, as prepared (Gd-doped) LTO/TiO2 didn’t show an obvious irreversible capacity. This is an important advantage over other anode materials in LIB. For example, Ndoped TiO2/graphene composites have shown excellent electrochemical performance, however, the irreversible capacity is still very large based on our recent study . The superior electrochemical performance for Gd-doped LTO/ TiO2 can be attributed to the synergetic effects of Gd doping and the presence of rutile TiO2 in which Gd doping has played a critical role in improving the electrochemical performance. One hand, the addition of Gd3+ leads to the formation of heteroatom-doped LTO, which is able to improve the intrinsic conductivity of LTO, lower the polarization and enhance the structural stabilization, resulting
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375
improved electrochemical performance [41–43]. It has been disclosed that the Gd doping can improve the electric conductivity as well as the ionic conductivity, help to enhancing the electrochemical performance based on DFT studies . On the other hand, the introduction of Gd3+ affects the growth behaviour of LTO nanosheets due to surface electric dipoles effects, leading to the formation of ultrathin morphology [6,44–46]. The thin nanosheets morphology for Gd-doped LTO/TiO2 help to the shorten diffusion length for both Li+ and electrons, which can reduce the concentration polarization and improve the rate performance, facilitating the charge-discharge process. The twodimensional ﬂat structure can also provide a larger surface contact area between nanosheets and conductive carbon black as well as the larger electrode-electrolyte contact area compared with the bulk material . Furthermore, the distortion in both LTO and rutile TiO2 caused by heteroatom doping might affect the band gap, leading to the change of ionic and electronic conductivity . It is noteworthy that the certain amount of rutile in the composites is beneﬁcial to improve the kinetics of the electrode due to the high Li-ion diffusion coefﬁcient for rutile TiO2 [35,47]. In addition, the introduction of rutile TiO2 will not decrease the speciﬁc capacity for the composites as rutile TiO2 can contribute very large speciﬁc capacity (theoretical capacity of 336 mAh/g). The electrochemical impedance spectroscopy was also used to evaluate the electrochemical performance of as prepared samples. As shown in Fig. 7, the Nyquist plots of all the three samples have shown the typical characteristics of one semicircle in middle-high frequency range and a sloping straight line in the low frequency. The semicircle in the middle frequency range indicated the charge transfer resistance, relating to the charge transfer through the electrode/electrolyte interface. As seen from the Nyquist plots, it can be found that the radius of the semicircle of Gd-doped LTO/TiO2 electrode is smallest, and that of commercial LTO is the largest among the three samples. The smaller the radius of semicircle is, the smaller the charge transfer resistance is. And the resistances were calculated on the base of ﬁtting circuit (inset of Fig. 7). The smallest semicircle of Gd-doped LTO/TiO2 electrode indicated its smallest charge transfer resistance (about 60.9 ohms) among these three electrodes. (204.6 and 333.9 ohms for undoped LTO/TiO2 and commercial LTO, respectively). This experiment demonstrated that the charge-transfer resistance of LTO electrode can be lowered through doping of Gdatoms and combination with rutile TiO2. Although the mechanism of effects of Gd-doping on LTO/TiO2 should be further studied such as using DFT simulation, these experiments have proved that as fabricated Gd-doped LTO/TiO2
electrode possessed superior electrochemical performance as anode material in LIB. 4. Conclusions In conclusion, single-crystalline Gd-doped Li4Ti5O12/rutile TiO2 nanosheets composites have been prepared via solvothermal synthesis assisted method. Through the combination of rutile coating and Gd-doping, the as prepared Gd-doped LTO/TiO2 composites have exhibited greatly improved electrochemical performances including the superior rate performance, excellent cycle life, lowered concentration polarization and decreased charge transfer resistance. At lower rates, both Gd-doped and undoped LTO/rutile TiO2 can deliver a speciﬁc capacity close to the theoretical value, while Gd-doped sample deliver a much higher value than undoped sample at higher rates. The improved electrochemical performances for Gd-doped LTO/TiO2 have been attributed to the following reasons. One hand, the introduction of Gd has led to the formation of the heteroatom doped LTO/TiO2 which improved their ionic conductivity. On the other hand, the ultrathin morphology for as prepared sample has shortened the diffusion length for both Li-ion and electrons. Furthermore, the diffusion coefﬁcient has been improved attributed to the presence of rutile TiO2 and the shortened diffusion length for Li ion and electrons due to the thin nanosheets morphology. Notably, the Gd doping together with Rutile TiO2 coating can avoid the irreversible capacity and the low tap density caused by carbon coating, which are be of great importance in practical application for electrode material. Considering the carbon-coating free nature for this technique, as prepared sample might be very promising in lithium ion batteries or Li-ion capacitor with high energy density. Acknowledgements This work was ﬁnancially supported by NSFC (Grant No. 51202212), Natural Science Foundation of Hebei province (Grant No. E2014203033). The authors also thank Dr. Wentao Hu for the helpful discussion on TEM analysis. References                    
Fig. 7. Electrochemical impedance spectra of the commercial LTO, as prepared LTO/ TiO2 and Gd-doped LTO/TiO2 electrodes (Inset shows the equivalent circuit).
M. Armand, J.-M. Tarascon, Nature 451 (2008) 652–657. J.-M. Tarascon, M. Armand, Nature 414 (2001) 359–367. S.-E. Lee, E. Kim, J. Cho, Electrochem. Solid-State Lett. 10 (2007) A1–A4. S. Liu, Z. Wang, C. Yu, H.B. Wu, G. Wang, Q. Dong, J. Qiu, A. Eychmüller, X.W. Lou, Adv. Mater. 25 (2013) 3462–3467. S. Goriparti, E. Miele, F. De Angelis, E. Di Fabrizio, R. Proietti Zaccaria, C. Capiglia, J. Power Sources 257 (2014) 421–443. G.B. Xu, W. Li, L.W. Yang, X.L. Wei, J.W. Ding, J.X. Zhong, P.K. Chu, J. Power Sources 276 (2015) 247–254. J.Z. Chen, L. Yang, S.H. Fang, Y.F. Tang, Electrochim. Acta 55 (2010) 6596–6600. B.F. Wang, J.S. Wang, J. Cao, H.H. Ge, Y.F. Tang, J. Power Sources 266 (2014) 150– 154. X.G. Hao, B.M. Bartlett, Adv. Energy Mater. 3 (2013) 753–761. S. Chen, Y. Xin, Y. Zhou, Y. Ma, H. Zhou, L. Qi, Energy Environ Sci. 7 (2014) 1924– 1930. C.F. Lin, X.Y. Fan, Y.L. Xin, F.Q. Cheng, M.O. Lai, H.H. Zhou, L. Lu, Nanoscale 6 (2014) 6651–6660. Y. Sha, B. Zhao, R. Ran, R. Cai, Z. Shao, J. Mater. Chem. A 1 (2013) 13233–13243. L.F. Shen, E. Uchaker, X.G. Zhang, G.Z. Cao, Adv. Mater. 24 (2012) 6502–6506. K. Song, D.H. Seo, M.R. Jo, Y.I. Kim, K. Kang, Y.M. Kang, J. Phys. Chem. Lett. 5 (2014) 1368–1373. G.D. Du, N. Sharma, V.K. Peterson, J.A. Kimpton, D.Z. Jia, Z.P. Guo, Adv. Funct. Mater. 21 (2011) 3990–3997. B.B. Tian, H.F. Xiang, L. Zhang, Z. Li, H.H. Wang, Electrochim Acta 55 (2010) 5453–5458. Y.Q. Ge, H. Jiang, K. Fu, C.H. Zhang, J.D. Zhu, C. Chen, Y. Lu, Y.P. Qiu, X.W. Zhang, J. Power Sources 272 (2014) 860–865. C.F. Lin, M.O. Lai, L. Lu, H.H. Zhou, Y.L. Xin, J. Power Sources 244 (2013) 272–279. C.C. Yang, H.C. Hu, S.J. Lin, W.C. Chien, J. Power Sources 258 (2014) 424–433. H.P. Liu, S.F. Bi, G.W. Wen, X.G. Teng, P. Gao, Z.J. Ni, Y.M. Zhu, F. Zhang, J. Alloy Compd. 543 (2012) 99–104. Y. Ma, B. Ding, G. Ji, J.Y. Lee, Acs Nano 7 (2013) 10870–10878.
Y. Li et al. / Electrochimica Acta 182 (2015) 368–375  H. Song, S.W. Yun, H.H. Chun, M.G. Kim, K.Y. Chung, H.S. Kim, B.W. Cho, Y.T. Kim, Energy Environ Sci. 5 (2012) 9903–9913.  B.A. Zhang, Y.S. Liu, Z.D. Huang, S. Oh, Y. Yu, Y.W. Mai, J.K. Kim, J. Mater. Chem. 22 (2012) 12133–12140.  Z.Q. Zhu, F.Y. Cheng, J. Chen, J. Mater. Chem. A 1 (2013) 9484–9490.  E. Pohjalainen, J. Kallioinen, T. Kallio, J. Power Sources 279 (2015) 481–486.  L. Zhao, Y.S. Hu, H. Li, Z.X. Wang, L.Q. Chen, Adv. Mater. 23 (2011) 1385–1388.  J.G. Kim, D.Q. Shi, M.S. Park, G. Jeong, Y.U. Heo, M. Seo, Y.J. Kim, J.H. Kim, S.X. Dou, Nano Res. 6 (2013) 365–372.  L. Cheng, X.L. Li, H.J. Liu, H.M. Xiong, P.W. Zhang, Y.Y. Xia, J. Electrochem Soc. 154 (2007) A692–A697.  Y. Li, Z. Wang, X.-J. Lv, J. Mater. Chem. A 2 (2014) 15473–15479.  Y. Li, X. Lv, J. Lu, J. Li, J. Phys. Chem. C 114 (2010) 21770–21774.  I.N. Kholmanov, C.W. Magnuson, A.E. Aliev, H.F. Li, B. Zhang, J.W. Suk, L.L. Zhang, E. Peng, S.H. Mousavi, A.B. Khanikaev, R. Piner, G. Shvets, R.S. Ruoff, Nano Lett. 12 (2012) 5679–5683.  X. Li, H. Wang, J.T. Robinson, H. Sanchez, G. Diankov, H. Dai, J. Am. Chem. Soc. 131 (2009) 15939–15944.  M. Krajewski, M. Michalska, B. Hamankiewicz, D. Ziolkowska, K.P. Korona, J.B. Jasinski, M. Kaminska, L. Lipinska, A. Czerwinski, J. Power Sources 245 (2014) 764–771.  M.M. Rahman, J.Z. Wang, M.F. Hassan, D. Wexler, H.K. Liu, Adv. Energy Mater. 1 (2011) 212–220.
 Y.-Q. Wang, L. Gu, Y.-G. Guo, H. Li, X.-Q. He, S. Tsukimoto, Y. Ikuhara, L.-J. Wan, J. Am. Chem. Soc. 134 (2012) 7874–7879.  Q. Zhang, M.G. Verde, J.K. Seo, X. Li, Y.S. Meng, J. Power Sources 280 (2015) 355–362.  L. Lutterotti, S. Matthies, H.-R. Wenk, A. Schultz, J. Richardson Jr., J. Appl. Phys. 81 (1997) 594–600.  Y. Shi, L. Wen, F. Li, H.-M. Cheng, J Power Sources 196 (2011) 8610–8617.  Y.-l. Li, N.-f. Chen, J.-p. Zhou, S.-l. Song, L.-f. liu, Z.-g. Yin, C.-l. Cai, J. Crystal Growth 265 (2004) 548–552.  Y. Yang, B. Qiao, X. Yang, L. Fang, C. Pan, W. Song, H. Hou, X. Ji, Adv. Func. Mater. 24 (2014) 4349–4356.  R. Cai, S. Jiang, X. Yu, B. Zhao, H. Wang, Z. Shao, J. Mater. Chem. 22 (2012) 8013–8021.  T.-F. Yi, Y. Xie, Q. Wu, H. Liu, L. Jiang, M. Ye, R. Zhu, J. Power Sources 214 (2012) 220–226.  Y.-J. Bai, C. Gong, N. Lun, Y.-X. Qi, J. Mater. Chem. A 1 (2013) 89–96.  C. Han, Y.-B. He, B. Li, H. Li, J. Ma, H. Du, X. Qin, Q.-H. Yang, F. Kang, ChemSusChem 7 (2014) 2567–2574.  S.X. Yu, L.W. Yang, Y.C. Li, X. Qi, X.L. Wei, J.X. Zhong, J. Crystal Growth 367 (2013) 62–67.  H. Qiu, G. Chen, R. Fan, C. Cheng, S. Hao, D. Chen, C. Yang, Chem. Commun. 47 (2011) 9648–9650.  M.L. Sushko, K.M. Rosso, J. Liu, J. Phys. Chem. C 114 (2010) 20277–20283.