Fabrication of CNTs/MnO₂ composite as a wrapping layer for surface modification of Cr-doped LiNi_0.5Mn_1.5O₄ for lithium ion batteries

Abstract

Needle-like MnO₂ is deposited on the surface of CNTs using a simple redox reaction and simultaneously forms a special wrapping layer on the surface and among the particles of Cr-doped LiNi_0.5Mn_1.5O₄ through an auto-grow process. The CNTs act as a conductive network of electrons and the needle-like MnO₂ increases the contact points among the particles and furthermore provides additional Li⁺ diffusion paths. The CNTs/MnO₂ composite is also a protective layer to suppress the side reaction between the electrode and the electrolyte. The CNTs/MnO₂-coated, Cr-doped LiNi_0.5Mn_1.5O₄ (CNTs/MnO₂–Cr-LNMO) shows superior electrochemical properties, with a high specific capacity and excellent cycling stability at high voltage and high rate.

---

Introduction

With the development of lithium ion batteries (LIBs), high-energy and high-power-density LIBs have proven to be a most promising technology.^1,2 Spinel LiNi_0.5Mn_1.5O₄ with three-dimensional (3D) Li⁺ diffusion channels is considered to be one of the best prospects among cathode materials, owing to its extremely high cyclic stability and relatively high specific capacity (146.7 mA h g⁻¹) with a plateau at around 4.7 V.^3,4 As is well known, there are two crystallographic structures of LiNi_0.5Mn_1.5O₄. A high-temperature synthesis usually produces a non-stoichiometric, disordered and oxygen-deficient LiNi_0.5Mn_1.5O_4−δ powder with a Fd3m structure, which shows better electrochemical performance than the ordered spinel with P4₃32, due to the effect of the oxygen deficiency in increasing the amount of Mn³⁺, and it manifests a higher Li⁺ diffusion coefficient, lower resistance and lattice strain during cycling.^5,6

However, oxygen release from the spinel also leads to the co-existence of Li_yNi_1−yO impurities, lowering the obtainable capacity.³ Furthermore, a portion of the Mn³⁺ can also form Mn²⁺ through a disproportionation reaction, and this results in significant capacity loss during cycling.^7,8 To eliminate the impurities and improve the cyclability, one commonly adopted approach is to partially substitute Ni and (or) Mn with other cations,^9–14 and Cr³⁺-doping is the most widely researched.^3,15,16

In addition, an interfacial side reaction between the cathode materials and the electrolyte is another issue hampering the adoption of LiNi_0.5Mn_1.5O₄. At a high operating voltage of 4.7 V, the carbonate-based liquid electrolyte can decompose and the production of HF will attack the cathode material, which triggers the dissolution of the transition metals and give rise to severe capacity fading.¹⁷ Consequently, researchers have devoted much attention to the introduction of metal oxides,^18,19 carbon materials²⁰ and Li⁺-conductors,^16,21 etc. as coating layers for LiNi_0.5Mn_1.5O₄, in order to decrease the contact area between the electrode and the electrolyte and further improve the cyclability.

Herein, we present a facile auto-grow process for depositing needle-like MnO₂ uniformly on the surface of carbon nanotubes (CNTs) through a simple redox reaction and for obtaining a special wrapping layer for Cr-doped LiNi_0.5Mn_1.5O₄. Through modifying the surfaces of the particles and their bulk phase, the CNTs/MnO₂-coated, Cr-doped LiNi_0.5Mn_1.5O₄ (CNTs/MnO₂–Cr-LNMO) showed superior electrochemical properties, with a high specific capacity and excellent cycling stability at a high rate. The effects of the CNTs/MnO₂ layer on the physical and electrochemical performance of LiNi_0.5Mn_1.5O₄ are discussed in detail.

Experimental

Sample preparation

LiNi_0.5Mn_1.5O₄ (LNMO) was prepared by a solid-state method. Stoichiometric amounts of LiOH, Ni(OH)₂ and MnO₂ were thoroughly mixed in an agate jar by ball milling at 400 rpm for 15 h, followed by a heat treatment at 900 °C for 12 h (with a 5 °C min⁻¹ cooling rate), and further re-annealing at 700 °C for 8 h (with a 1 °C min⁻¹ cooling rate) in air. For the sample LiNi_0.45Cr_0.05Mn_1.5O₄ (Cr-LNMO), an appropriate amount of Cr(NO₃)₃·9H₂O was added to the starting materials. To compensate for the loss of Li during treatment at high temperature, an additional 5% of LiOH starting material was used in all samples.

The CNTs/MnO₂ composite was prepared by a redox reaction between acid-treated CNTs and KMnO₄. First of all, the CNTs were treated with a H₂SO₄/HNO₃ (3 : 1 volume ratio) solution at 70 °C for 8 h. A mixed aqueous suspension containing the acid-treated CNTs (33 mg) and Cr-LNMO (326 mg) was then subjected to ultrasonic vibration for 30 min. Then, 50 mL of 20 mM KMnO₄ aqueous solution was added into the suspension dropwise; meanwhile the suspension was heated at 70 °C in air for 2 h, and the dark purple liquid became deep brown during the process. The precipitation was centrifuged, washed several times and dried in air at 60 °C overnight. Finally, the powders were annealed at 350 °C for 1 h. The whole experimental design is summarized in Scheme 1.

Scheme 1 Schematic illustration of the experimental design.

Structural and morphological characterization

Powder X-ray diffraction (XRD) measurements were performed on a Rigaku Dmax2500 diffractometer, with Cu Kα radiation between 10° and 90°. The thermal analysis (TG-DTA) study was performed using a HENVEN HCT-1/2 instrument, at a scan rate of 10 °C min⁻¹ from 25 °C to 700 °C. The Fourier transform infrared spectrum (FT-IR) of the product was obtained by use of a Magna 560 spectrometer (American Nicolet), using the KBr pellet technique. The morphology and microstructure of the products were studied using a field emission scanning electron microscope (SEM, Hitachi SU8010) equipped with an energy-dispersive X-ray detector for elemental analysis and mapping, and also with a high-resolution transmission electron microscope (HRTEM, JEOL-2100F). The cyclic voltammograms (CV) of the cells were acquired on a P4000 electrochemical workstation, at a scan rate of 0.1 mV s⁻¹ between 3.0–5.1 V and at various rates between 3.5–5.0 V. The electrochemical impedance spectroscopy (EIS) tests were carried out using a P4000 workstation in the range 0.01 Hz to 100 kHz, at room temperature.

Electrochemical measurements

Electrochemical measurements were carried out using a coin-type cell (CR2025) with metallic lithium as counter electrodes. The cathode electrodes were fabricated by dispersing 70 wt% active materials, 20 wt% 350 carbon and 10 wt% poly(vinylidene fluoride) (PVDF) binder in N-methyl-2-pyrrolidone (NMP) solvent. The homogeneous slurry was then painted onto aluminum foil, followed by drying at 120 °C in a vacuum oven overnight. The loading of the active material on the electrode was 2.0 ± 0.5 mg cm⁻², and the capacities were calculated based on the total weight of the active materials and the coating. All cells were assembled in an Ar-filled glove box. Cellgard 2400 was used as separators and the electrolyte was 1 M LiPF₆ dissolved in ethylene carbonate (EC) and dimethyl carbonate (DMC) (1 : 1 by volume). The charge–discharge tests were performed on a LAND CT2001A battery testing system in the voltage range from 3.5–5.0 V versus Li/Li⁺ at room temperature.

Results and discussion

Fig. 1a shows the X-ray diffraction patterns (XRD) of the as-prepared LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO powders. There are no obvious peaks of Li_yNi_1−yO impurity detected by XRD, which means that the impurity is absent or the content is quite low. The re-annealing process can reduce the oxygen deficiency, and thus decrease the formation of Li_yNi_1−yO and produce a pure product.³ The sharp peaks indicate that all the samples were well crystallized. There are no obvious differences in XRD patterns of the three samples. However, the lattice parameter is increased after Cr³⁺ doping (Table 1), which is due to the appearance of Mn³⁺ in the samples. When the Ni(2+) is substituted partially by the higher valence Cr(3+), part of the Mn⁴⁺ (0.53 Å) is accordingly reduced to the larger Mn³⁺ (0.65 Å), on account of electrical neutrality. The CNTs/MnO₂-coated samples show little difference in their crystal structure compared to Cr-LNMO. There are no peaks related to MnO₂ or CNTs in the patterns, due to the small amount of the coating layer, as well as the detection limit of XRD. Thus, thermal analysis (TG-DTA) was undertaken to provide evidence of the wrapping layer. The results (Fig. 1b) confirm that the main weight loss of 3.89% from 350 °C to 550 °C is attributable to the oxidation of CNTs and the oxygen loss from the lattice of MnO₂.²²

Fig. 1 (a) X-ray diffraction patterns of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO; (b) TG-DTA curves of CNTs/MnO₂–Cr-LNMO from 100 to 700 °C.

Table 1 The physical parameters, specific capacity and coulombic efficiency for the 1^st charge–discharge cycle at 0.2C

Sample	Space group	a	k	D(Li^+)	Initial discharge capacity	Initial coulombic efficiency
LNMO	P4332	8.17(3) Å	—	—	123.2 mA h g^−1	72.83%
Cr-LNMO	Fd3m	8.18(6) Å	0.0191	4.9 × 10^−12 cm^2 s^−1	129.5 mA h g^−1	73.07%
MnO2/CNTs–Cr-LNMO	Fd3m	8.18(2) Å	0.0402	2.2 × 10^−11 cm^2 s^−1	146.4 mA h g^−1	83.34%

---

Infrared and Raman spectroscopy have been proven to be valid ways to distinguish between Fd3m and P4₃32 crystal structures. The FTIR spectra of the samples are given in Fig. 2a. According to the work of Kunduraci and Amatucci,²³ the dominant stretching vibrations are at ∼592 cm⁻¹ (Ni–O) and ∼621 cm⁻¹ (Mn–O). The spectrum of LNMO is consistent with the ordered phase (P4₃32) on account of a larger ratio of the two bands at 592 cm⁻¹ and 621 cm⁻¹. Furthermore, the bands around 430, 465, 555 and 651 cm⁻¹ are characteristic peaks of the ordered phase, which are almost absent in the spectrum of the other two cation-disordered (Fd3m) samples. This indicates that the pristine LNMO sample is dominated by the Ni/Mn ordered phase in the bulk,⁶ while the Cr-LNMO and CNTs/MnO₂–Cr-LNMO powders display the typical profiles of a cubic structure with the space group Fd3m. The change in the dominant phase in the spinel is due to the content of Mn³⁺ in the samples being tuned by Cr³⁺ doping.

Fig. 2 (a) The Fourier transformed infrared (FTIR) and (b) Raman spectra of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO.

The Raman spectra are shown in Fig. 2b. The strong band around 635 cm⁻¹ assigned to the symmetric Mn–O stretching vibration of the MnO₆ octahedra (A_1g) and the features at 412 cm⁻¹ and 492 cm⁻¹ assigned to the Ni–O stretching mode are obvious in the LNMO sample, which are all consistent with the ordered phase (P4₃32). On the other hand, no characteristic peaks of the ordered structure of P4₃32, namely 240, 220 and 160 cm⁻¹, are detected in the spectrum of Cr-LNMO and CNTs/MnO₂–Cr-LNMO.²² The result of the Raman investigation is in agreement with the FTIR spectra. In addition, two weak bands around 1590 and 1350 cm⁻¹ are the G band and D band of carbon, shown in the inset graph. The G band represents the in-plane bond-stretching motion of the pairs of C sp² atoms, whereas the D band represents the breathing modes of rings or K-point phonons of A_1g symmetry.²⁴

The morphologies of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO are compared in Fig. 3. It can be seen that the particles of all samples (Fig. 3a–c) show the same octahedral shape, with smooth and clean surface facets. After the coating treatment, the spinel particles still retain a well-crystallized morphology. It can be seen from Fig. 3c and d that the Cr-LNMO octahedral particles are embedded uniformly in a network composed of CNTs/MnO₂ nanotubes, forming a close wrapping layer coating on the surface of the sample. In other words, a CNTs/MnO₂ nanotube net has grown on the surface of the particles in situ, which can be seen clearly from the transmission electron microscope (TEM) analysis (Fig. 4). The HRTEM image in Fig. 4b illustrates the interplanar spacing, measured to be 0.47 nm, which is the lattice distance of LNMO, corresponding to d(111) calculated from XRD. In the enlarged map of the CNTs/MnO₂ nanotube net shown in Fig. 4c, the needle-like MnO₂ are extensively distributed on the surface of the nanotubes. This was enlarged further to observe the interplanar spacing of the tube wall and the needle-like branch on the surface of the tube from the HRTEM image (Fig. 4d). The interplanar spacing of the tube wall is 0.34 nm, related to the d(002) plane of CNTs, and that of the branch on the surface of the tube is 0.23 nm, related to the d(100) plane of MnO₂. The ultra-thin MnO₂ is deposited on the surface of 1D CNTs through a simple redox reaction:

4MnO₄⁻ + 3C + H₂O = 4MnO₂ + CO₃²⁻ + 2HCO₃⁻ (1)

and simultaneously forms a special wrapping layer for Cr-LNMO during an auto-grow process. In this structure, the CNTs can act as a conducting matrix to build a bridge between electrons and Cr-LNMO, and the MnO₂ will improve the transport rate of Li⁺.^25,26 This wrapping layer not only covers the surfaces of the Cr-LNMO particles, but also penetrates the bulk material of the particles, reinforcing the contact between interfaces and also facilitating the absorption of electrolyte. Hence it influences the Li⁺ transfer and the formation of SEI. In addition, it can protect the active materials by suppressing the dissolution of transition metals. From this point of view, the CNTs/MnO₂ modified Cr-LNMO electrode should exhibit excellent capacity and cycle stability at a high operating voltage.

Fig. 3 SEM images of (a) LNMO; (b) Cr-LNMO; (c and d) CNTs/MnO₂–Cr-LNMO.

Fig. 4 (a and c) TEM and (b and d) HRTEM images of CNTs/MnO₂–Cr-LNMO.

In order to confirm the elemental composition of CNTs/MnO₂–Cr-LNMO, the SEM and corresponding EDS mapping images are given in Fig. 5, which demonstrates that the elements of O, Mn, Ni, and C are well distributed in the sample. All of the elements have clear boundaries except for Cr, which may because the small amount of Cr cannot be detected precisely.

Fig. 5 SEM and corresponding EDS elemental mapping images of CNTs/MnO₂–Cr-LNMO.

To understand the differences in the Li insertion/extraction behaviors of the samples, cyclic voltammograms (CV) and galvanostatic charge–discharge curves of cells were characterized. It can be seen from Fig. 6a that there is only a single redox peak for the sample LNMO at around 4.7 V, owning to the redox reaction of Ni^2+/4+. In contrast, the 4.0 V peaks corresponding to the Mn^3+/4+ redox couple appear in the curves of Cr-LNMO and CNTs/MnO₂–Cr-LNMO (Fig. 6c and e), meaning there is a small content of Mn³⁺ in the samples, which is also confirmed in the charge–discharge profiles (Fig. 6d and f). Moreover, the redox peaks of Ni at 4.7 V have split into two peaks in the samples of Cr-LNMO and CNTs/MnO₂–Cr-LNMO, and the peak positions shift slightly to a lower voltage. It is known that Li⁺ requires more energy to be intercalated/deintercalated from the host structure in the higher-voltage region in the case of LNMO.⁵ The voltage shift to the lower-voltage region could be ascribed to the formation of a different structure. It can also be seen from the charge–discharge curves that two separate plateaus appear at ∼4.0 V and 4.7 V owing to the Mn^3+/4+ and Ni^2+/3+/4+, which correspond to the oxidation peaks of the CV curves. The initial discharge capacity and coulombic efficiency of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO are listed in Table 1, The CNTs/MnO₂–Cr-LNMO sample shows the highest specific capacity of 146.4 mA h g⁻¹ and the best coulombic efficiency of 83.34%, which is owing to the excellent conductivity and kinetics of lithium diffusion of the CNTs/MnO₂–Cr-LNMO.

Fig. 6 (a, c and e) CV curves at a scan rate of 0.1 mV s⁻¹ between 3.0 and 5.1 V and (b, d and f) initial charge–discharge profiles of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO at 0.2C.

Lithium ion migration rate was estimated by a series of CV tests as a function of the scan rate (ν) in the range from 0.1 to 1.0 mV s⁻¹ (Fig. 7). The variation in peak shape reflects the kinetics of lithium intercalation/deintercalation at the electrode/electrolyte interface. Assuming that the intercalation/deintercalation reaction is controlled by the solid-state diffusion of Li⁺, then the dependence of i_p on ν^1/2 can be applied to determine the diffusion coefficient of Li⁺ (D_Li) on the basis of the following equation:^16,27,28

i_p = 2.69 × 10⁵n^3/2AD_Li^1/2ν^1/2C_Li (2)

where i_p is the peak current, n is the number of electrons per reaction species, A is the total surface area of the electrode (in this case the value is 1.33 cm²), and C_Li is the concentration of Li⁺ in the electrode (given as 0.024 mol cm⁻³). Based on eqn (2), the calculated D_Li values of Cr-LNMO and CNTs/MnO₂–Cr-LNMO are 4.9 × 10⁻¹² and 2.2 × 10⁻¹¹ cm² s⁻¹, respectively (Table 1), which demonstrates that the kinetics of Li⁺ diffusion is obviously improved by modification of the CNTs/MnO₂ wrapping layer.

Fig. 7 (a and b) CV curves of Cr-LNMO and CNTs/MnO₂–Cr-LNMO cells at different scan rates from 0.1 to 1.0 mV s⁻¹ and (c) plots of peak current (i_p) of the cyclic voltammograms vs. ν^1/2.

The cycle performances of three samples at 1C rate are shown in Fig. 8a. The capacity of LNMO decreased severely after the 300th cycle. By contrast, the capacity retention of Cr-LNMO and CNTs/MnO₂–Cr-LNMO can reach 83.9% and 93.1% after 500 cycles, respectively. For CNTs/MnO₂–Cr-LNMO, the capacity increased in the first few cycles, and reached a maximum value of 136.2 mA h g⁻¹. We list relevant working values of Cr-doped LNMO in Table 2, and their performances are comparable. The CNTs/MnO₂–Cr-LNMO sample also exhibits excellent rate performances, as shown in Fig. 8b. The discharge capacities at different rates remain good from 0.2C to 5C and the platform is still holding at about 4.6 V with the increase in the rates. Even at 10C, the capacity can still maintain ∼80 mA h g⁻¹ with a high voltage of ∼4.5 V. We attribute the excellent cycle performance to the wrapping layer of CNTs/MnO₂, which acts as a protective layer to suppress the metal dissolution and helps to reduce the side reactions, preventing the structure from being destroyed by the electrolyte.

Fig. 8 (a) Cycling performance of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO at 1C. (b) Discharge profiles of CNTs/MnO₂–Cr-LNMO at various rates between 3.5–5.0 V.

Table 2 Comparison of the electrochemical performance for the Cr-doped LNMO in the literature and in this paper

No.	Material	Synthesis method	Active material loading	Current rate	Discharge capacity		Ref.
					1st	cycles
1	LiNi0.45Cr0.05Mn1.5O4	Solid state method	—	0.1C	141 mA h g^−1	—	15
2	LiNi0.45Cr0.1Mn1.45O4	Thermopolymerization method	3.0–3.5 mg cm^−2	1C	140 mA h g^−1	114 mA h g^−1 (500)	29
3	LiNi0.3Cr0.2Mn1.5O4	Co-precipitation method	—	0.2 mA cm^−1	138 mA h g^−1	—	30
4	LiNi0.42Cr0.08Mn1.5O4	Co-precipitation method	—	10 mA g^−1	132.5 mA h g^−1	—	31
5	LiNi0.45Cr0.1Mn1.45O4	PVP-combusting method	3.2 mg cm^−2	1C	126 mA h g^−1	120 mA h g^−1 (500)	32
6	LiNi0.45Cr0.05Mn1.5O4	Solid state method	5.0–10 mg cm^−2	1C	120 mA h g^−1	119 mA h g^−1 (250)	3
7	LiNi0.45Cr0.1Mn1.45O4	Post-annealing method	—	0.2C	115 mA h g^−1	113 mA h g^−1 (100)	33
8	CNTs/MnO2–Cr-LNMO	Solid state method	2.0 ± 0.5 mg cm^−2	1C	127.5 mA h g^−1	126.3 mA h g^−1 (500)	This work

---

In order to further confirm this, we conducted CV tests to compare the differences between Cr-LNMO and CNTs/MnO₂–Cr-LNMO after long-term cycling. As shown in Fig. 9, the peaks of the redox couples moved slightly towards lower voltages for both samples after charging/discharging at 10C for 100 cycles, corresponding to the degradation of the electrode. However, CNTs/MnO₂–Cr-LNMO shows a smaller voltage difference of cathodic and anodic peaks, which indicates its better reaction reversibility and a lower electrode polarization. Thus, the CNTs/MnO₂–Cr-LNMO sample has a more stable structure after long-term cycling due to the suppression of metal dissolution, which is also consistent with the cycle and rate results.

Fig. 9 CV curves of Cr-LNMO and CNTs/MnO₂–Cr-LNMO after charging and discharging at 10C for 100 cycles.

Electrochemical impedance spectroscopy (EIS) analysis was taken before and after 500 cycles at a stable open-circuit voltage. The equivalent circuit model used to fit the impedance spectra is shown in Fig. 10c. As described in the literature,^15,16 a resistor of R_sf and a capacitor of C_sf connected in parallel are added to the circuit to take into account the formation of surface film at the electrode surface, and the intermediate frequency area consists of a charge transfer resistance (R_ct) and a double layer capacitance (C_dL) of the electrode-electrolyte interface. At low frequency, a Warburg impedance (W_d) arises from the diffusion of Li⁺ in the bulk materials. The electrolyte resistance R₀ is quite low and the values remain almost constant, which demonstrates the low variation of electrolyte concentration and conductivity. Apparently, the resistance of the cells dramatically increased after 500 cycles in the order of LNMO > Cr-LNMO > CNTs/MnO₂–Cr-LNMO. The cell of CNTs/MnO₂–Cr-LNMO shows the lowest impedance, owing to the high diffusion capability of Li ions and electrons through the cathode and SEI films.

Fig. 10 AC impedance of LNMO, Cr-LNMO and CNTs/MnO₂–Cr-LNMO at a stable open-circuit voltage (a) before and (b) after 500 cycles. (c) Equivalent circuit for fitting experimental EIS data.

Conclusion

In summary, we successfully synthesized a high-voltage spinel cathode material LiNi_0.45Cr_0.05Mn_1.5O₄ with CNTs/MnO₂ nanotube modification through a solid-state method combined with a facile auto-grow redox reaction. The CNTs/MnO₂ network simultaneously wound among the Cr-LNMO particles and coated the surface of the sample, providing excellent electron and Li⁺ diffusion paths. Moreover, the CNTs/MnO₂ wrapping layer also acts as a protective film on the surface of the materials to suppress the metal dissolution in the electrolyte and further keep the structure stability of the sample. The CNTs/MnO₂–Cr-LNMO cathode exhibits a high initial discharge capacity of 146.4 mA h g⁻¹ and coulombic efficiency of 83.34% at 0.2C. The capacity retention of CNTs/MnO₂–Cr-LNMO is up to 93.1% after 500 cycles even at a 1C rate. It can be determined that this design of materials with a CNTs/MnO₂ wrapping among the particles and bulk materials is an efficient way to improve the electrochemical performance of electrode materials.

Acknowledgements

Support for this work was provided by Changbai Mountain Scholar Project by Education Department of Jilin Province.

Notes and references

1. A. Manthiram, K. Chemelewski and E.-S. Lee, Energy Environ. Sci., 2014, 7, 1339–1350.
2. A. Kraytsberg and Y. Ein-Eli, Adv. Energy Mater., 2012, 2, 922–939.
3. J. Xiao, X. Chen, P. V. Sushko, M. L. Sushko, L. Kovarik, J. Feng, Z. Deng, J. Zheng, G. L. Graff, Z. Nie, D. Choi, J. Liu, J. G. Zhang and M. S. Whittingham, Adv. Mater., 2012, 24, 2109–2116.
4. T.-F. Yi, J. Mei and Y.-R. Zhu, J. Power Sources, 2016, 316, 85–105.
5. J. H. Kim, S. T. Myung, C. S. Yoon, S. G. Kang and Y. K. Sun, Chem. Mater., 2004, 16, 906–914.
6. X. Zhang, J. Liu, H. Yu, G. Yang, J. Wang, Z. Yu, H. Xie and R. Wang, Electrochim. Acta, 2010, 55, 2414–2417.
7. N. P. W. Pieczonka, Z. Liu, P. Lu, K. L. Olson, J. Moote, B. R. Powell and J.-H. Kim, J. Phys. Chem. C, 2013, 117, 15947–15957.
8. R. Qiao, Y. Wang, P. Olalde-Velasco, H. Li, Y.-S. Hu and W. Yang, J. Power Sources, 2015, 273, 1120–1126.
9. J. Liu and A. Manthiram, J. Phys. Chem. C, 2009, 113, 15073–15079.
10. M.-H. Liu, H.-T. Huang, C.-M. Lin, J.-M. Chen and S.-C. Liao, Electrochim. Acta, 2014, 120, 133–139.
11. M. Mo, K. S. Hui, X. Hong, J. Guo, C. Ye, A. Li, N. Hu, Z. Huang, J. Jiang, J. Liang and H. Chen, Appl. Surf. Sci., 2014, 290, 412–418.
12. W. Zhu, D. Liu, J. Trottier, C. Gagnon, A. Guerfi, C. M. Julien, A. Mauger and K. Zaghib, J. Power Sources, 2014, 264, 290–298.
13. G. H. Lee, H. S. Kim, S. G. Baek, H. J. Choi, K. Y. Chung, B. W. Cho, S. Y. Lee and Y.-S. Lee, J. Power Sources, 2015, 298, 379–384.
14. L. Wang, D. Chen, J. Wang, G. Liu, W. Wu and G. Liang, RSC Adv., 2015, 5, 99856–99865.
15. Y.-P. Zeng, X.-l. Wu, P. Mei, L.-N. Cong, C. Yao, R.-S. Wang, H.-M. Xie and L.-Q. Sun, Electrochim. Acta, 2014, 138, 493–500.
16. X. Yang, T. Yang, S. Liang, X. Wu and H. Zhang, J. Mater. Chem. A, 2014, 2, 10359.
17. L. Yang, B. Ravdel and B. L. Lucht, Electrochem. Solid-State Lett., 2010, 13, A95.
18. J. Chong, S. Xun, J. Zhang, X. Song, H. Xie, V. Battaglia and R. Wang, Chemistry, 2014, 20, 7479–7485.
19. J. W. Kim, D. H. Kim, D. Y. Oh, H. Lee, J. H. Kim, J. H. Lee and Y. S. Jung, J. Power Sources, 2015, 274, 1254–1262.
20. X. Fang, M. Ge, J. Rong and C. Zhou, J. Mater. Chem. A, 2013, 1, 4083.
21. H. Deng, P. Nie, H. Luo, Y. Zhang, J. Wang and X. Zhang, J. Mater. Chem. A, 2014, 2, 18256–18262.
22. H. Pang, Z. Yang, J. Lv, W. Yan and T. Guo, Energy, 2014, 69, 392–398.
23. M. Kunduraci and G. G. Amatucci, J. Electrochem. Soc., 2006, 153, A1345.
24. J. Yan, Z. Fan, T. Wei, W. Qian, M. Zhang and F. Wei, Carbon, 2010, 48, 3825–3833.
25. W. Wei, X. Cui, W. Chen and D. G. Ivey, Chem. Soc. Rev., 2011, 40, 1697–1721.
26. A. L. M. Reddy, M. M. Shaijumon, S. R. Gowda and P. M. Ajayan, Nano Lett., 2009, 9, 1002–1006.
27. X. Zhang, F. Cheng, K. Zhang, Y. Liang, S. Yang, J. Liang and J. Chen, RSC Adv., 2012, 2, 5669.
28. T. Yang, N. Zhang, Y. Lang and K. Sun, Electrochim. Acta, 2011, 56, 4058–4064.
29. G. B. Zhong, Y. Y. Wang, Y. Q. Yu and C. H. Chen, J. Power Sources, 2012, 205, 385–393.
30. T. A. Arunkumar and A. Manthiram, Electrochim. Acta, 2005, 50, 5568–5572.
31. E.-S. Lee and A. Manthiram, J. Mater. Chem. A, 2013, 1, 3118.
32. J. Mao, K. Dai, M. Xuan, G. Shao, R. Qiao, W. Yang, V. S. Battaglia and G. Liu, ACS Appl. Mater. Interfaces, 2016, 8, 9116–9124.
33. D. Liu, J. Hamel-Paquet, J. Trottier, F. Barray, V. Gariépy, P. Hovington, A. Guerfi, A. Mauger, C. M. Julien, J. B. Goodenough and K. Zaghib, J. Power Sources, 2012, 217, 400–406.

---