- Email: [email protected]
NaTi2(PO4)3 rechargeable aqueous sodium-ion battery
Journal of Power Sources 327 (2016) 327e332
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Electrolyte dependence of the performance of a Na2FeP2O7// NaTi2(PO4)3 rechargeable aqueous sodium-ion battery Kosuke Nakamoto a, Yusuke Kano a, Ayuko Kitajou b, Shigeto Okada b, * a b
Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-Koen, Kasuga, Fukuoka 816-8580, Japan
h i g h l i g h t s Na2FeP2O7 with aqueous electrolytes has higher rate capability than non-aqueous. Na2FeP2O7 with concentrated electrolytes has larger capacity than diluted. NaTi2(PO4)3 with NaNO3 aq. shows irreversiblility with corrosive side reactions.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 April 2016 Received in revised form 14 June 2016 Accepted 16 July 2016
Aqueous sodium-ion battery is attractive, because of the low cost and the high safety. However, since the electrochemical window of the aqueous electrolyte is narrow, there have been few reports concerning the optimum cathode materials for use in aqueous sodium-ion batteries up to now. This work focused on Na2FeP2O7 as a cathode material for a novel aqueous sodium-ion battery, and investigated the electrolyte dependence of the performances of a Na2FeP2O7//NaTi2(PO4)3 full-cell. The battery performances such as the rate capability and cyclability of Na2FeP2O7//NaTi2(PO4)3 full-cell with 2 M Na2SO4 or 4 M NaClO4 aqueous electrolyte were better than that with the non-aqueous electrolyte. However, a Na2FeP2O7// NaTi2(PO4)3 full-cell with 4 M NaNO3 aqueous electrolyte exhibited a large irreversible capacity due to the corrosive side reaction. © 2016 Published by Elsevier B.V.
Keywords: Sodium-ion battery Aqueous electrolyte Sodium iron pyrophosphate Sodium titanium phosphate
1. Introduction To achieve the efficient utilization of the renewable energies, such as solar and wind power, the large-scale grid energy storage for load leveling system is an indispensable device. At present, there are Ni-Cd battery, Ni-metal hydride (MH) battery, Li-ion battery and sodium-sulfur battery as the commercially available rechargeable batteries. Among them, the Li-ion battery has certain advantages, such as high voltage and high energy density. However, the performance needs to be improved with regard to battery cycle life, production costs and safety. In particular, high-quality lithium provided by the sun drying of the salt lake is relatively costly, and the annual production in the world is insufficient. Na-ion batteries, which have a working mechanism similar to that of Li-ion batteries, have attracted much interest as a power source for large-scale grid energy storage, because of the low cost and abundance of sodium.
* Corresponding author. E-mail address: [email protected] (S. Okada). http://dx.doi.org/10.1016/j.jpowsour.2016.07.052 0378-7753/© 2016 Published by Elsevier B.V.
However, both Li- and Na-ion batteries have an associated safety issue: at elevated temperatures, charged oxide cathodes such as LixCoO2 readily release oxygen gas, which can cause the exothermic reaction with the organic electrolytes [1,2]. In addition, conventional organic electrolytes are low conductive and fairly expensive. To overcome these drawbacks, aqueous Na-ion batteries with noninflammable, high conductive and inexpensive aqueous electrolytes rather than organic electrolytes have been proposed. Recently, Aquion Energy Ltd. released a commercially-available energy storage system composed of a Na4Mn9O18 (Na0.44MnO2) cathode, activated carbon (AC) anode and Na2SO4 aqueous electrolyte [3,4]. This Na0.44MnO2//AC system is not a rocking-chair type cell, but a hybrid capacitor. That is, the cathode reaction involves the intercalation/deintercalation of Naþ ions into/from the Na0.44MnO2 bronze matrix, while the anode reaction consists of the adsorption/desorption of SO2 on the surface of AC. Since the 4 rechargeable capacity of this hybrid capacitor is restricted by the concentration of SO2 4 anion in the electrolyte, the rechargeable capacity per weight or volume is smaller than that of a rocking-
328
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
chair type cell. Although several cathode active materials for the aqueous Naion battery have been proposed in recent years [5e16], reports of the anode active materials are still limited. Among them, minormetal free NASICON-type NaTi2(PO4)3 (NTP) is a rare and valuable anode example for the aqueous Na-ion battery [5,6]. The flat charge/discharge profile, corresponding to Ti3þ/Ti4þ redox reaction, locates within the electrochemical window of water. As the most inexpensive 3d transition metal, iron-based cathode active materials have also been proposed. In particular, olivine-type NaFePO4, as the Na counterpart of LiFePO4 is promising. However, it cannot be obtained by the direct synthesis method, because electrochemically inactive maricite-type NaFePO4 is more stable than olivine-type NaFePO4. Therefore, the only synthetic route to olivine-type NaFePO4 is ion exchange between Naþ and Liþ in LiFePO4 [17,18]. Because the ion-exchange process is relatively costly, concerted efforts are being devoted to finding new low-cost materials and manufacturing processes. Our own group has focused on Na2FeP2O7 as a promising cathode material for aqueous Na-ion batteries, because Na2FeP2O7 can be obtained by simple, conventional solid state processes [19e23]. This compound exhibits a reversible specific capacity of approximately 57 mAh g1 with an average voltage of 0.05 V vs. SCE, and therefore has a redox potential within the electrochemical window of water [22]. In the present work, we evaluated the electrochemical performance of Na2FeP2O7 as a cathode for aqueous Na-ion batteries, in conjunction with typical Na-salt electrolytes, such as 0.5e4 M Na2SO4, NaNO3 and NaClO4 solutions. In addition, we investigated the electrochemical performance of an aqueous Na2FeP2O7//NaTi2(PO4)3 full-cell with various Na salt concentrations. 2. Experimental 2.1. Materials synthesis As cathode active material, Na2FeP2O7 was synthesized according to a previously reported procedure [19]. Briefly, stoichiometric amounts of NaH2PO4 and Fe(COO)2$2H2O (Wako Pure Chemical Industries, Ltd.) were placed in an alumina container along with 3 mmf zirconia balls. The precursor was subsequently prepared by wet ball-milling in acetone at 400 rpm for 2 h. After evaporating the acetone to obtain a dry mass, the precursor was ground in an agate mortar and pressed into cylindrical pellets. These pellets were sintered at 600 C for 10 h under Ar containing 5% H2. After cooling to ambient temperature, the desired Na2FeP2O7 powder (as-prepared Na2FeP2O7) was obtained. On the other hand, anode active material, NaTi2(PO4)3 was prepared from a stoichiometric mixture of Na2CO3 (Kishida Chemical Co., Ltd.), titanium (IV) butoxide (Sigma-Aldrich Co. LLC.) and NH4H2PO4 (Nacalai Tesque, INC.) using the Pechini method [5,24]. This precursor mixture was decomposed at 350 C for 3 h in air to eliminate the ammonia and organic components. The resulting powder was ground, pressed into cylindrical pellets and calcined at 800 C for 12 h in air. The products were characterized by X-ray powder diffraction (XRD, 50 kV and 300 mA, Cu-Ka, RINT 2100HLR/PC, Rigaku Corporation) while the chemical composition of the Na2FeP2O7 was ascertained using inductively coupled plasma-atomic emission spectroscopy (ICP-AES; PerkinElmer Optima 8300). In ICP measurement, the powder was dissolved in a mixture of 30 wt% H2O2 and conc. aqueous HCl. 2.2. Electrochemical properties To improve the electronic conductivity, 25 wt% acetylene black
(AB) was subsequently added to 70 wt% Na2FeP2O7 powder and the mixtures were dry ball-milled at 300 rpm for 10 h. In addition to the carbon coating, the carbothermal annealing for the obtained Na2FeP2O7/C composite was performed at 600 C for 10 h under Ar containing 5% H2 [21]. Hereinafter, the Na2FeP2O7/C carbon composites after carbon coating and carbothermal annealing are referred to as CC and CT samples, respectively. In combination with 5 wt% polytetrafluoroethylene binder (Polyflon PTFE F-104, Daikin Industries, Ltd.), the cathode pellets were fabricated from the CC and CT carbon composites (95 wt%), respectively. On the other hand, 70 wt% NaTi2(PO4)3 anode powder was combined with 25 wt% AB and ball-milled at 400 rpm for 1 h. Then, for the obtained NaTi2(PO4)3/C composite, the carbothermal treatment was done at 800 C for 1 h under N2 [5]. To fabricate the anode pellet, 95 wt% NaTi2(PO4)3/C carbon composite was fabricated from with 5 wt% PTFE binder. They were subsequently punched into discs (10 mmf). These discs were sandwiched between sheets of nickel mesh (Thank Metal Co., LTD.). A three-electrode electrochemical cells incorporating various aqueous electrolytes were used in cyclic voltammetry and galvanostatic discharge/charge tests. An Ag-AgCl electrode with saturated KCl (RE-6, BAS Inc.) and pure zinc (99.9%, Nilaco Corp.) were used as the reference and counter electrodes, respectively. The electrochemical performances of the Na2FeP2O7// NaTi2(PO4)3 full-cell with various aqueous electrolytes were evaluated by a 2032 coin-type cell setup. The cathode/anode mass balance in this full-cell was 2:3 and the cathode/anode capacity balance was approximately 1:3. Employed aqueous electrolytes consisted of 0.5 M Na2SO4, 2 M Na2SO4, 1 M NaNO3, 4 M NaNO3, 1 M NaClO4 and 4 M NaClO4, respectively. Dissolved oxygen was purged from each aqueous electrolyte by bubbling Ar, since the Na3Ti2(PO4)3 generated on charging process is readily oxidized by dissolved oxygen in an aqueous solution [25]. The electrochemical performance of the Na2FeP2O7 was also confirmed in a 2032 cointype cell with non-aqueous electrolyte (1 M NaClO4 in propylene carbonate (PC), Tomiyama Pure Chemical Industries, LTD.) and a polypropylene separator and Na metal (Sigma-Aldrich Co. LLC.). The cell incorporating the non-aqueous electrolyte and Na metal was assembled in an Ar-filled glove box (dew point < 80 C). Cyclic voltammetry was performed with a Versastat 3 (AMETEK, Inc.). Galvanostatic charge/discharge tests were carried out using a cycler (NAGANO & Co., Ltd.) at a constant current density. 3. Result and discussion 3.1. Structural characterization of Na2FeP2O7 The obtained samples were characterized by XRD, generating the patterns presented in Fig. 1. The Na2FeP2O7 powder was identified as triclinic with P-1 diffraction pattern consistent with Na1.56Fe1.22P2O7 (ICDD # 83e0255), with no impurity phases detectable. The XRD profile of the CC sample (Fig. 1 (b)) contained peaks that were broader than those of the as-prepared Na2FeP2O7, indicating the lower crystallinity. In contrast, the XRD peaks of the CT sample became sharper following the carbothermal treatment, suggesting that the crystallinity of the Na2FeP2O7 was improved by annealing at 600 C. 3.2. Electrochemical performances of a Na2FeP2O7 cathode against Zn counter electrode Fig. 2 compares the initial and second charge-discharge profiles of CC and CT cathodes in 2 M Na2SO4 aqueous electrolyte. These electrochemical measurements were carried out within a potential
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
Fig. 1. XRD patterns obtained from (a) as-prepared Na2FeP2O7, (b) carbon-coated Na2FeP2O7 (CC sample) and (c) carbothermally annealed Na2FeP2O7 (CT sample).
329
the same 1 M Naþ concentration, at a rate of 2.0 mA cm2. These electrochemical measurements were carried out within a potential range from 0.6 to 0.7 V (vs. Ag/AgCl in aqueous electrolyte) and from 2.3 to 3.6 V (vs. Na/Naþ in 1 M NaClO4/PC). The initial charge and discharge capacities for Na2FeP2O7 with 1 M NaClO4/PC were 53 and 50 mAh g1, while the values obtained with 0.5 M Na2SO4, 1 M NaNO3 and 1 M NaClO4 aqueous electrolytes were 28/44, 29/47 and 27/46 mAh g1, respectively. The polarization between the discharge and charge voltage profiles in employing aqueous electrolytes was smaller than that observed in the non-aqueous electrolyte. Because the viscosity of the water is lower than that of nonaqueous solvent. In Fig. 3 (b), as the Naþ concentration thickened, the initial charge/discharge capacities of Na2FeP2O7 were increased. This improvement of the cathode performances are well explained by the increasing the ionic conductivity along with increasing salt concentration in these solutions as shown in Table 1. The chemical compositions of the Na2FeP2O7 products were assessed using ICP-AES, and the formula Na1.68FeP2O7 was determined. Na2FeP2O7 is known to be easy to be oxidized by Na release in air [19]. Therefore, Naþ ions are believed to insert into empty sites in the Na1.68FeP2O7 during the initial discharge process. Fig. 4 (a) summarizes the cyclabilities of Na2FeP2O7 with various electrolytes at a rate of 2.0 mA cm2. The initial discharge capacities of the Na2FeP2O7 were 50 mAh g1 (1 M NaClO4 in PC), 58 mAh g1 (2 M Na2SO4 in aq.), 69 mAh g1 (4 M NaNO3 in aq.) and 68 mAh g1 (4 M NaClO4 in aq.). The capacity retention values, which are the ratio of the discharge capacity at 30th cycle for the discharge capacity at 1st cycle, were 71% (1 M NaClO4 in PC), 89% (2 M Na2SO4 in aq.), 92% (4 M NaNO3 in aq.) and 96% (4 M NaClO4 in aq.), respectively. From these results, Na2FeP2O7 exhibits relatively good cyclability, regardless of the electrolytes salt in PC. The rate capabilities of the Na2FeP2O7 with various electrolytes were also shown in Fig. 4 (b). The initial discharge capacity of Na2FeP2O7 with 4 M NaNO3 or 4 M NaClO4 aqueous electrolyte remained above 60 mAh g1 even at a rate of 5.0 mA cm2, although the capacity obtained in 1 M NaClO4 in PC was only 10 mAh g1. Ionic conductivity of non-aqueous Naþ electrolyte, such as NaClO4 in PC decreases in high concentrated region more than about 1 M, because of the high viscosity [26]. On the other hand, ionic conductivity of aqueous electrolyte, such as NaClO4 aq., increases until about 6 M [10]. Water can dissolve much salt due to its higher relative dielectric constant than non-aqueous solvent, and high concentrate aqueous electrolyte is higher ion conductivity than non-aqueous electrolyte. These common positive results for aqueous cells with high Naþ concentration bring the good battery performances. 3.3. Electrochemical performance of a Na2FeP2O7//NaTi2(PO4)3 fullcell
Fig. 2. The initial and second charge/discharge curves of the CC and CT samples using 2 M Na2SO4 aq. as the electrolyte with a potential window from 0.6 to 0.7 V (vs. Ag/ AgCl).
range from 0.6 to 0.7 V (vs. Ag/AgCl) at a rate of 0.2 mA cm2. The initial charge/discharge capacities were 60/68 mAh g1 for CC cathode and 59/78 mAh g1 for CT cathode, respectively. Especially, the CT cathode shows the smaller polarization in Fig. 2 due to the higher crystallinity. Fig. 3 (a) presents the initial charge-discharge curves obtained from the Na2FeP2O7 in conjunction with various electrolytes having
Fig. 5 (a) summarizes the initial and second charge/discharge curves of the Na2FeP2O7//NaTi2(PO4)3 full-cell with various electrolytes at a rate of 2.0 mA cm2. The initial charge and discharge capacities of this full-cell with 1 M NaClO4 in PC were 50 and 47 mAh g1, with a 94% discharge/charge efficiency. The discharge/ charge efficiency values obtained in conjunction with 2 M Na2SO4, 4 M NaNO3 and 4 M NaClO4 in aq. were 95, 82 and 96%, respectively. As a result, although the Na2FeP2O7 half-cell with 4 M NaNO3 aq. electrolyte showed a small irreversible capacity and good cyclability, the irreversible capacity of Na2FeP2O7//NaTi2(PO4)3 with 4 M NaNO3 aq. electrolyte was not small. To investigate the reason, cyclic voltammograms with a NaTi2(PO4)3 anode using various aqueous electrolytes were performed (Fig. 5 (b)). The cathodic and anodic peak pair located at 0.91 V (red.) and 0.68 V (ox.) vs. Ag/ AgCl was observed in 2 M Na2SO4 and at 0.82 V (red.) and 0.68 V
330
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
Fig. 3. The initial and second charge/discharge curves of the CT sample using aqueous electrolytes containing (a) 1 M and (b) 4 M Naþ ions.
Table 1 Relationship between the ionic conductivities of the electrolytes and the obtained capacities of the Na2FeP2O7 cathode in half-cell (Fig. 3). Electrolytes 1 M NaClO4 in PC 0.5 M Na2SO4 aq. 1 M NaClO4 aq. 1 M NaNO3 aq. 2 M Na2SO4 aq. 4 M NaNO3 aq. 4 M NaClO4 aq.
Ionic conductivity /mS cm1
Na2FeP2O7 discharge capacity /mAh g1
5.8 [10] 65 [26] 75 [26] 80 [26] 125 [26] 165 [26] 165 [26]
50 44 46 47 64 69 68
Fig. 4. (a) Cyclabilities and (b) rate capabilities of the CT sample using various electrolytes.
(ox.) in 4 M NaClO4 aq. electrolyte, respectively. They suggest the reversible Ti4þ/Ti3þ redox reaction of NaTi2(PO4)3 anode in the aqueous electrolytes [5]. However, in the case of the 4 M NaNO3 aq. electrolyte, the reduction peak below 1.15 V corresponding to H2 gas generation by the reduction
of water was larger than those in the other aqueous electrolytes. In addition, irreversible behaviors were observed around 1.1 V on the first and second anodic sweep. It suggests the deterioration of anode material through a corrosion reaction by nitric acid. Nitric acid anion will be reduced to form nitrous acid anion as follows;
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
331
Fig. 5. (a) The initial and second charge/discharge curves of the Na2FeP2O7//NaTi2(PO4)3 full-cell using various electrolytes and (b) the cycle profiles of cyclic voltammograms obtained from NaTi2(PO4)3 anode materials using 2 M Na2SO4, 4 M NaNO3 and 4 M NaClO4 in aq. electrolyte.
Fig. 6. (a) Cyclabilities and (b) rate capabilities of the Na2FeP2O7//NaTi2(PO4)3 full-cell using various electrolytes. e 2NO 3 þ 2H2O þ 2e / N2O4 þ 4OH
(1)
According to the potential estimation of Gibbs' free energy (E ¼ eDG/nF), the NO 3 reduction of Chemical eq. (1) occurs below 1.04 V vs. Ag/AgCl. So, the relatively large anodic current below 1.04 V in the 4 M NaNO3 aqueous electrolyte in Fig. 5 (b) is reasonable. Then, HNO2 is immediately generated in water as Chemical eq. (2). N2O4 þ H2O / HNO2 þ HNO3
(2)
HNO2 oxidant generated at low voltage region should reduce the charged Na3Ti2(PO4)3 anode in the cell. This is a reasonable explanation of the decreased discharge capacity in 4 M NaNO3 aqueous electrolyte in Fig. 5 (a). Actually, the similar degradation behavior of anode side has been reported even in the aqueous Liion battery (LiMn2O4/5 M LiNO3 aq./LiTi2(PO4)3) [27]. Actually, the capacity retentions of Na2FeP2O7//NaTi2(PO4)3 full-cells with 2 M Na2SO4 and 4 M NaClO4 aq. electrolyte were relatively high and almost similar to that of 1 M NaClO4 in PC as shown in Fig. 6 (a).
However, the Na2FeP2O7//NaTi2(PO4)3 full-cell with 4 M NaNO3 retained very little of its reversible capacity after 30 cycles. Fig. 6 (b) compares the rate capabilities of the Na2FeP2O7//NaTi2(PO4)3 fullcells with various electrolytes. The cells with aqueous electrolytes exhibit improved rate capability compared to that of the nonaqueous electrolyte. In particular, very little rechargeable capacity is obtained with 1 M NaClO4 in PC at a rate of 5.0 mA cm2. These results suggest that the high ionic conductivity of aqueous electrolytes can assists the rapid insertion/extraction of the large Naþ ions into/from the host matrices of the cathode and anode active materials.
4. Conclusion We investigated the electrochemical performances of Na2FeP2O7//NaTi2(PO4)3 full-cell with various aqueous electrolytes. The aqueous electrolyte has some advantages such as nonflammable, high ionic conductivity and low costs compared to nonaqueous electrolyte. In particular, the electrochemical
332
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
performances (including rate capability, cyclability and polarization between discharge and charge voltage profiles) of the Na2FeP2O7// NaTi2(PO4)3 full-cell with 2 M Na2SO4 or 4 M NaClO4 were improved relative to those obtained with the non-aqueous electrolyte. This is attributed to the higher relative dielectric constant and lower viscosity of water than organic solvent. Conversely, the Na2FeP2O7//NaTi2(PO4)3 full-cell with 4 M NaNO3 exhibited a large irreversible capacity due to H2 gas evolution and corrosive side reactions. On the other hand, it may be difficult for the NaClO4 salts to use in market due to their explosiveness and oxidizing abilities. We therefore propose a Na2FeP2O7//NaTi2(PO4)3 full-cell in conjunction with 2 M Na2SO4 as a promising aqueous sodium-ion battery candidate. Acknowledgements This work was financially supported by Elements Strategy Initiative for Catalysts and Batteries Project, MEXT, Japan. References [1] P.G. Balakrishnan, R. Ramesh, T.P. Kumar, J. Power Sources 155 (2006) 401e414. [2] J. Zhao, L. Zhao, N. Dimov, S. Okada, T. Nishida, J. Electrochem. Soc. 160 (6) (2013) A3077eA3081. [3] A.D. Tevar, J.F. Whitacre, J. Electrochem. Soc. 157 (7) (2010) A870eA875. [4] J.F. Whitacre, T. Wiley, S. Shanbhag, Y. Wenzhuo, A. Mohamed, S.E. Chun, E. Weber, D. Blackwood, E. Lynch-Bell, J. Gulakowski, C. Smith, D. Humphreys, J. Power Sources 213 (2012) 255e264. [5] S.I. Park, I. Gocheva, S. Okada, J. Yamaki, J. Electrochem. Soc. 158 (10) (2011) A1067eA1070. [6] W. Wu, A. Mohamed, J.F. Whitacre, J. Electrochem. Soc. 160 (3) (2013)
A497eA504. [7] Z. Li, D.B. Ravnsbæk, K. Xiang, Y.M. Chiang, Electrochem. Commun. 44 (2014) 12e15. [8] H. Qin, Z.P. Song, H. Zhan, Y.H. Zhou, J. Power Sources 249 (2014) 367e372. [9] Z. Li, D. Young, K. Xiang, W.C. Carter, Y.M. Chiang, Adv. Energy Mater. 3 (2013) 290e294. [10] W. Wu, S. Shabhag, J. Chang, A. Rutt, J.F. Whitacre, J. Electrochem. Soc. 162 (6) (2015) A803eA808. [11] X. Wu, Y. Cao, X. Ai, J. Qian, H. Yang, Electrochem. Commun. 31 (2013) 145e148. [12] C.D. Wessells, S.V. Peddada, R.A. Huggins, Y. Cui, Nano Lett. 11 (2011) 5421e5425. [13] X.Y. Wu, M.Y. Sun, Y.F. Shen, J.F. Qian, Y.L. Cao, X.P. Ai, H.X. Yang, Chem. Sus. Chem. 7 (2014) 407e411. [14] W. Song, X. Ji, Y. Zhu, H. Zhu, F. Li, J. Chen, F. Lu, Y. Yao, C.E. Banks, Chem. Electro. Chem. 1 (2014) 871e876. [15] C. Deng, S. Zhang, Z. Dong, Y. Shang, Nano Energy 4 (2014) 49e55. [16] Y. Liu, B.H. Zhang, S.Y. Xiao, L.L. Liu, Z.B. Wen, Y.P. Wu, Electrochim. Acta 116 (2014) 512e517. [17] S.M. Oh, S.T. Myung, J. Hassoun, B. Scrosati, Y.K. Sun, Electrochem. Commun. 22 (2012) 149e152. [18] M. Vujkovi c, S. Mentus, J. Power Sources 247 (2014) 184e188. [19] P. Barpanda, G. Liu, C.D. Ling, M. Tamaru, M. Avdeev, S.C. Chung, Y. Yamada, A. Yamada, Chem. Mater 25 (2013) 3480e3487. [20] T. Honma, T. Togashi, N. Ito, T. Komatsu, J. Ceram. Soc. Jpn. 120 (8) (2012) 344e346. [21] H. Kim, R.A. Shakoor, C. Park, S.Y. Lim, J.S. Kim, Y.N. Jo, W. Cho, K. Miyasaka, R. Kahraman, Y. Jung, J.W. Choi, Adv. Funct. Mater. 23 (2013) 1147e1155. [22] Y.H. Jung, C.H. Lim, J.H. Kim, D.K. Kim, RSC Adv. 4 (2014) 9799e9802. [23] C.Y. Chen, K. Matsumoto, T. Nohira, R. Hagiwara, J. Electrochem. Soc. 162 (1) (2015) A176eA180. [24] C.R. Mariappan, C. Galven, M.-P. Crosnier-Lopez, F. Le Berre, O. Bohnke, J. Solid State Chem. 179 (2006) 450e456. [25] J.Y. Luo, W.J. Cui, P. He, Y.Y. Xia, Nat. Chem. 2 (2010) 760e765. [26] K. Kuratani, I. Kishimoto, Y. Nishida, R. Kondo, H.T. Takeshita, H. Senoh, T. Kiyobayashi, J. Electrochem. Soc. 163 (2016) H417eH425. [27] H. Wang, K. Huang, Y. Zeng, S. Yang, L. Chen, Electrochem. Acta 52 (2007) 3280e3285.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Electrolyte dependence of the performance of a Na2FeP2O7// NaTi2(PO4)3 rechargeable aqueous sodium-ion battery Kosuke Nakamoto a, Yusuke Kano a, Ayuko Kitajou b, Shigeto Okada b, * a b
Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-Koen, Kasuga, Fukuoka 816-8580, Japan
h i g h l i g h t s Na2FeP2O7 with aqueous electrolytes has higher rate capability than non-aqueous. Na2FeP2O7 with concentrated electrolytes has larger capacity than diluted. NaTi2(PO4)3 with NaNO3 aq. shows irreversiblility with corrosive side reactions.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 April 2016 Received in revised form 14 June 2016 Accepted 16 July 2016
Aqueous sodium-ion battery is attractive, because of the low cost and the high safety. However, since the electrochemical window of the aqueous electrolyte is narrow, there have been few reports concerning the optimum cathode materials for use in aqueous sodium-ion batteries up to now. This work focused on Na2FeP2O7 as a cathode material for a novel aqueous sodium-ion battery, and investigated the electrolyte dependence of the performances of a Na2FeP2O7//NaTi2(PO4)3 full-cell. The battery performances such as the rate capability and cyclability of Na2FeP2O7//NaTi2(PO4)3 full-cell with 2 M Na2SO4 or 4 M NaClO4 aqueous electrolyte were better than that with the non-aqueous electrolyte. However, a Na2FeP2O7// NaTi2(PO4)3 full-cell with 4 M NaNO3 aqueous electrolyte exhibited a large irreversible capacity due to the corrosive side reaction. © 2016 Published by Elsevier B.V.
Keywords: Sodium-ion battery Aqueous electrolyte Sodium iron pyrophosphate Sodium titanium phosphate
1. Introduction To achieve the efficient utilization of the renewable energies, such as solar and wind power, the large-scale grid energy storage for load leveling system is an indispensable device. At present, there are Ni-Cd battery, Ni-metal hydride (MH) battery, Li-ion battery and sodium-sulfur battery as the commercially available rechargeable batteries. Among them, the Li-ion battery has certain advantages, such as high voltage and high energy density. However, the performance needs to be improved with regard to battery cycle life, production costs and safety. In particular, high-quality lithium provided by the sun drying of the salt lake is relatively costly, and the annual production in the world is insufficient. Na-ion batteries, which have a working mechanism similar to that of Li-ion batteries, have attracted much interest as a power source for large-scale grid energy storage, because of the low cost and abundance of sodium.
* Corresponding author. E-mail address: [email protected] (S. Okada). http://dx.doi.org/10.1016/j.jpowsour.2016.07.052 0378-7753/© 2016 Published by Elsevier B.V.
However, both Li- and Na-ion batteries have an associated safety issue: at elevated temperatures, charged oxide cathodes such as LixCoO2 readily release oxygen gas, which can cause the exothermic reaction with the organic electrolytes [1,2]. In addition, conventional organic electrolytes are low conductive and fairly expensive. To overcome these drawbacks, aqueous Na-ion batteries with noninflammable, high conductive and inexpensive aqueous electrolytes rather than organic electrolytes have been proposed. Recently, Aquion Energy Ltd. released a commercially-available energy storage system composed of a Na4Mn9O18 (Na0.44MnO2) cathode, activated carbon (AC) anode and Na2SO4 aqueous electrolyte [3,4]. This Na0.44MnO2//AC system is not a rocking-chair type cell, but a hybrid capacitor. That is, the cathode reaction involves the intercalation/deintercalation of Naþ ions into/from the Na0.44MnO2 bronze matrix, while the anode reaction consists of the adsorption/desorption of SO2 on the surface of AC. Since the 4 rechargeable capacity of this hybrid capacitor is restricted by the concentration of SO2 4 anion in the electrolyte, the rechargeable capacity per weight or volume is smaller than that of a rocking-
328
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
chair type cell. Although several cathode active materials for the aqueous Naion battery have been proposed in recent years [5e16], reports of the anode active materials are still limited. Among them, minormetal free NASICON-type NaTi2(PO4)3 (NTP) is a rare and valuable anode example for the aqueous Na-ion battery [5,6]. The flat charge/discharge profile, corresponding to Ti3þ/Ti4þ redox reaction, locates within the electrochemical window of water. As the most inexpensive 3d transition metal, iron-based cathode active materials have also been proposed. In particular, olivine-type NaFePO4, as the Na counterpart of LiFePO4 is promising. However, it cannot be obtained by the direct synthesis method, because electrochemically inactive maricite-type NaFePO4 is more stable than olivine-type NaFePO4. Therefore, the only synthetic route to olivine-type NaFePO4 is ion exchange between Naþ and Liþ in LiFePO4 [17,18]. Because the ion-exchange process is relatively costly, concerted efforts are being devoted to finding new low-cost materials and manufacturing processes. Our own group has focused on Na2FeP2O7 as a promising cathode material for aqueous Na-ion batteries, because Na2FeP2O7 can be obtained by simple, conventional solid state processes [19e23]. This compound exhibits a reversible specific capacity of approximately 57 mAh g1 with an average voltage of 0.05 V vs. SCE, and therefore has a redox potential within the electrochemical window of water [22]. In the present work, we evaluated the electrochemical performance of Na2FeP2O7 as a cathode for aqueous Na-ion batteries, in conjunction with typical Na-salt electrolytes, such as 0.5e4 M Na2SO4, NaNO3 and NaClO4 solutions. In addition, we investigated the electrochemical performance of an aqueous Na2FeP2O7//NaTi2(PO4)3 full-cell with various Na salt concentrations. 2. Experimental 2.1. Materials synthesis As cathode active material, Na2FeP2O7 was synthesized according to a previously reported procedure [19]. Briefly, stoichiometric amounts of NaH2PO4 and Fe(COO)2$2H2O (Wako Pure Chemical Industries, Ltd.) were placed in an alumina container along with 3 mmf zirconia balls. The precursor was subsequently prepared by wet ball-milling in acetone at 400 rpm for 2 h. After evaporating the acetone to obtain a dry mass, the precursor was ground in an agate mortar and pressed into cylindrical pellets. These pellets were sintered at 600 C for 10 h under Ar containing 5% H2. After cooling to ambient temperature, the desired Na2FeP2O7 powder (as-prepared Na2FeP2O7) was obtained. On the other hand, anode active material, NaTi2(PO4)3 was prepared from a stoichiometric mixture of Na2CO3 (Kishida Chemical Co., Ltd.), titanium (IV) butoxide (Sigma-Aldrich Co. LLC.) and NH4H2PO4 (Nacalai Tesque, INC.) using the Pechini method [5,24]. This precursor mixture was decomposed at 350 C for 3 h in air to eliminate the ammonia and organic components. The resulting powder was ground, pressed into cylindrical pellets and calcined at 800 C for 12 h in air. The products were characterized by X-ray powder diffraction (XRD, 50 kV and 300 mA, Cu-Ka, RINT 2100HLR/PC, Rigaku Corporation) while the chemical composition of the Na2FeP2O7 was ascertained using inductively coupled plasma-atomic emission spectroscopy (ICP-AES; PerkinElmer Optima 8300). In ICP measurement, the powder was dissolved in a mixture of 30 wt% H2O2 and conc. aqueous HCl. 2.2. Electrochemical properties To improve the electronic conductivity, 25 wt% acetylene black
(AB) was subsequently added to 70 wt% Na2FeP2O7 powder and the mixtures were dry ball-milled at 300 rpm for 10 h. In addition to the carbon coating, the carbothermal annealing for the obtained Na2FeP2O7/C composite was performed at 600 C for 10 h under Ar containing 5% H2 [21]. Hereinafter, the Na2FeP2O7/C carbon composites after carbon coating and carbothermal annealing are referred to as CC and CT samples, respectively. In combination with 5 wt% polytetrafluoroethylene binder (Polyflon PTFE F-104, Daikin Industries, Ltd.), the cathode pellets were fabricated from the CC and CT carbon composites (95 wt%), respectively. On the other hand, 70 wt% NaTi2(PO4)3 anode powder was combined with 25 wt% AB and ball-milled at 400 rpm for 1 h. Then, for the obtained NaTi2(PO4)3/C composite, the carbothermal treatment was done at 800 C for 1 h under N2 [5]. To fabricate the anode pellet, 95 wt% NaTi2(PO4)3/C carbon composite was fabricated from with 5 wt% PTFE binder. They were subsequently punched into discs (10 mmf). These discs were sandwiched between sheets of nickel mesh (Thank Metal Co., LTD.). A three-electrode electrochemical cells incorporating various aqueous electrolytes were used in cyclic voltammetry and galvanostatic discharge/charge tests. An Ag-AgCl electrode with saturated KCl (RE-6, BAS Inc.) and pure zinc (99.9%, Nilaco Corp.) were used as the reference and counter electrodes, respectively. The electrochemical performances of the Na2FeP2O7// NaTi2(PO4)3 full-cell with various aqueous electrolytes were evaluated by a 2032 coin-type cell setup. The cathode/anode mass balance in this full-cell was 2:3 and the cathode/anode capacity balance was approximately 1:3. Employed aqueous electrolytes consisted of 0.5 M Na2SO4, 2 M Na2SO4, 1 M NaNO3, 4 M NaNO3, 1 M NaClO4 and 4 M NaClO4, respectively. Dissolved oxygen was purged from each aqueous electrolyte by bubbling Ar, since the Na3Ti2(PO4)3 generated on charging process is readily oxidized by dissolved oxygen in an aqueous solution [25]. The electrochemical performance of the Na2FeP2O7 was also confirmed in a 2032 cointype cell with non-aqueous electrolyte (1 M NaClO4 in propylene carbonate (PC), Tomiyama Pure Chemical Industries, LTD.) and a polypropylene separator and Na metal (Sigma-Aldrich Co. LLC.). The cell incorporating the non-aqueous electrolyte and Na metal was assembled in an Ar-filled glove box (dew point < 80 C). Cyclic voltammetry was performed with a Versastat 3 (AMETEK, Inc.). Galvanostatic charge/discharge tests were carried out using a cycler (NAGANO & Co., Ltd.) at a constant current density. 3. Result and discussion 3.1. Structural characterization of Na2FeP2O7 The obtained samples were characterized by XRD, generating the patterns presented in Fig. 1. The Na2FeP2O7 powder was identified as triclinic with P-1 diffraction pattern consistent with Na1.56Fe1.22P2O7 (ICDD # 83e0255), with no impurity phases detectable. The XRD profile of the CC sample (Fig. 1 (b)) contained peaks that were broader than those of the as-prepared Na2FeP2O7, indicating the lower crystallinity. In contrast, the XRD peaks of the CT sample became sharper following the carbothermal treatment, suggesting that the crystallinity of the Na2FeP2O7 was improved by annealing at 600 C. 3.2. Electrochemical performances of a Na2FeP2O7 cathode against Zn counter electrode Fig. 2 compares the initial and second charge-discharge profiles of CC and CT cathodes in 2 M Na2SO4 aqueous electrolyte. These electrochemical measurements were carried out within a potential
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
Fig. 1. XRD patterns obtained from (a) as-prepared Na2FeP2O7, (b) carbon-coated Na2FeP2O7 (CC sample) and (c) carbothermally annealed Na2FeP2O7 (CT sample).
329
the same 1 M Naþ concentration, at a rate of 2.0 mA cm2. These electrochemical measurements were carried out within a potential range from 0.6 to 0.7 V (vs. Ag/AgCl in aqueous electrolyte) and from 2.3 to 3.6 V (vs. Na/Naþ in 1 M NaClO4/PC). The initial charge and discharge capacities for Na2FeP2O7 with 1 M NaClO4/PC were 53 and 50 mAh g1, while the values obtained with 0.5 M Na2SO4, 1 M NaNO3 and 1 M NaClO4 aqueous electrolytes were 28/44, 29/47 and 27/46 mAh g1, respectively. The polarization between the discharge and charge voltage profiles in employing aqueous electrolytes was smaller than that observed in the non-aqueous electrolyte. Because the viscosity of the water is lower than that of nonaqueous solvent. In Fig. 3 (b), as the Naþ concentration thickened, the initial charge/discharge capacities of Na2FeP2O7 were increased. This improvement of the cathode performances are well explained by the increasing the ionic conductivity along with increasing salt concentration in these solutions as shown in Table 1. The chemical compositions of the Na2FeP2O7 products were assessed using ICP-AES, and the formula Na1.68FeP2O7 was determined. Na2FeP2O7 is known to be easy to be oxidized by Na release in air [19]. Therefore, Naþ ions are believed to insert into empty sites in the Na1.68FeP2O7 during the initial discharge process. Fig. 4 (a) summarizes the cyclabilities of Na2FeP2O7 with various electrolytes at a rate of 2.0 mA cm2. The initial discharge capacities of the Na2FeP2O7 were 50 mAh g1 (1 M NaClO4 in PC), 58 mAh g1 (2 M Na2SO4 in aq.), 69 mAh g1 (4 M NaNO3 in aq.) and 68 mAh g1 (4 M NaClO4 in aq.). The capacity retention values, which are the ratio of the discharge capacity at 30th cycle for the discharge capacity at 1st cycle, were 71% (1 M NaClO4 in PC), 89% (2 M Na2SO4 in aq.), 92% (4 M NaNO3 in aq.) and 96% (4 M NaClO4 in aq.), respectively. From these results, Na2FeP2O7 exhibits relatively good cyclability, regardless of the electrolytes salt in PC. The rate capabilities of the Na2FeP2O7 with various electrolytes were also shown in Fig. 4 (b). The initial discharge capacity of Na2FeP2O7 with 4 M NaNO3 or 4 M NaClO4 aqueous electrolyte remained above 60 mAh g1 even at a rate of 5.0 mA cm2, although the capacity obtained in 1 M NaClO4 in PC was only 10 mAh g1. Ionic conductivity of non-aqueous Naþ electrolyte, such as NaClO4 in PC decreases in high concentrated region more than about 1 M, because of the high viscosity [26]. On the other hand, ionic conductivity of aqueous electrolyte, such as NaClO4 aq., increases until about 6 M [10]. Water can dissolve much salt due to its higher relative dielectric constant than non-aqueous solvent, and high concentrate aqueous electrolyte is higher ion conductivity than non-aqueous electrolyte. These common positive results for aqueous cells with high Naþ concentration bring the good battery performances. 3.3. Electrochemical performance of a Na2FeP2O7//NaTi2(PO4)3 fullcell
Fig. 2. The initial and second charge/discharge curves of the CC and CT samples using 2 M Na2SO4 aq. as the electrolyte with a potential window from 0.6 to 0.7 V (vs. Ag/ AgCl).
range from 0.6 to 0.7 V (vs. Ag/AgCl) at a rate of 0.2 mA cm2. The initial charge/discharge capacities were 60/68 mAh g1 for CC cathode and 59/78 mAh g1 for CT cathode, respectively. Especially, the CT cathode shows the smaller polarization in Fig. 2 due to the higher crystallinity. Fig. 3 (a) presents the initial charge-discharge curves obtained from the Na2FeP2O7 in conjunction with various electrolytes having
Fig. 5 (a) summarizes the initial and second charge/discharge curves of the Na2FeP2O7//NaTi2(PO4)3 full-cell with various electrolytes at a rate of 2.0 mA cm2. The initial charge and discharge capacities of this full-cell with 1 M NaClO4 in PC were 50 and 47 mAh g1, with a 94% discharge/charge efficiency. The discharge/ charge efficiency values obtained in conjunction with 2 M Na2SO4, 4 M NaNO3 and 4 M NaClO4 in aq. were 95, 82 and 96%, respectively. As a result, although the Na2FeP2O7 half-cell with 4 M NaNO3 aq. electrolyte showed a small irreversible capacity and good cyclability, the irreversible capacity of Na2FeP2O7//NaTi2(PO4)3 with 4 M NaNO3 aq. electrolyte was not small. To investigate the reason, cyclic voltammograms with a NaTi2(PO4)3 anode using various aqueous electrolytes were performed (Fig. 5 (b)). The cathodic and anodic peak pair located at 0.91 V (red.) and 0.68 V (ox.) vs. Ag/ AgCl was observed in 2 M Na2SO4 and at 0.82 V (red.) and 0.68 V
330
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
Fig. 3. The initial and second charge/discharge curves of the CT sample using aqueous electrolytes containing (a) 1 M and (b) 4 M Naþ ions.
Table 1 Relationship between the ionic conductivities of the electrolytes and the obtained capacities of the Na2FeP2O7 cathode in half-cell (Fig. 3). Electrolytes 1 M NaClO4 in PC 0.5 M Na2SO4 aq. 1 M NaClO4 aq. 1 M NaNO3 aq. 2 M Na2SO4 aq. 4 M NaNO3 aq. 4 M NaClO4 aq.
Ionic conductivity /mS cm1
Na2FeP2O7 discharge capacity /mAh g1
5.8 [10] 65 [26] 75 [26] 80 [26] 125 [26] 165 [26] 165 [26]
50 44 46 47 64 69 68
Fig. 4. (a) Cyclabilities and (b) rate capabilities of the CT sample using various electrolytes.
(ox.) in 4 M NaClO4 aq. electrolyte, respectively. They suggest the reversible Ti4þ/Ti3þ redox reaction of NaTi2(PO4)3 anode in the aqueous electrolytes [5]. However, in the case of the 4 M NaNO3 aq. electrolyte, the reduction peak below 1.15 V corresponding to H2 gas generation by the reduction
of water was larger than those in the other aqueous electrolytes. In addition, irreversible behaviors were observed around 1.1 V on the first and second anodic sweep. It suggests the deterioration of anode material through a corrosion reaction by nitric acid. Nitric acid anion will be reduced to form nitrous acid anion as follows;
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
331
Fig. 5. (a) The initial and second charge/discharge curves of the Na2FeP2O7//NaTi2(PO4)3 full-cell using various electrolytes and (b) the cycle profiles of cyclic voltammograms obtained from NaTi2(PO4)3 anode materials using 2 M Na2SO4, 4 M NaNO3 and 4 M NaClO4 in aq. electrolyte.
Fig. 6. (a) Cyclabilities and (b) rate capabilities of the Na2FeP2O7//NaTi2(PO4)3 full-cell using various electrolytes. e 2NO 3 þ 2H2O þ 2e / N2O4 þ 4OH
(1)
According to the potential estimation of Gibbs' free energy (E ¼ eDG/nF), the NO 3 reduction of Chemical eq. (1) occurs below 1.04 V vs. Ag/AgCl. So, the relatively large anodic current below 1.04 V in the 4 M NaNO3 aqueous electrolyte in Fig. 5 (b) is reasonable. Then, HNO2 is immediately generated in water as Chemical eq. (2). N2O4 þ H2O / HNO2 þ HNO3
(2)
HNO2 oxidant generated at low voltage region should reduce the charged Na3Ti2(PO4)3 anode in the cell. This is a reasonable explanation of the decreased discharge capacity in 4 M NaNO3 aqueous electrolyte in Fig. 5 (a). Actually, the similar degradation behavior of anode side has been reported even in the aqueous Liion battery (LiMn2O4/5 M LiNO3 aq./LiTi2(PO4)3) [27]. Actually, the capacity retentions of Na2FeP2O7//NaTi2(PO4)3 full-cells with 2 M Na2SO4 and 4 M NaClO4 aq. electrolyte were relatively high and almost similar to that of 1 M NaClO4 in PC as shown in Fig. 6 (a).
However, the Na2FeP2O7//NaTi2(PO4)3 full-cell with 4 M NaNO3 retained very little of its reversible capacity after 30 cycles. Fig. 6 (b) compares the rate capabilities of the Na2FeP2O7//NaTi2(PO4)3 fullcells with various electrolytes. The cells with aqueous electrolytes exhibit improved rate capability compared to that of the nonaqueous electrolyte. In particular, very little rechargeable capacity is obtained with 1 M NaClO4 in PC at a rate of 5.0 mA cm2. These results suggest that the high ionic conductivity of aqueous electrolytes can assists the rapid insertion/extraction of the large Naþ ions into/from the host matrices of the cathode and anode active materials.
4. Conclusion We investigated the electrochemical performances of Na2FeP2O7//NaTi2(PO4)3 full-cell with various aqueous electrolytes. The aqueous electrolyte has some advantages such as nonflammable, high ionic conductivity and low costs compared to nonaqueous electrolyte. In particular, the electrochemical
332
K. Nakamoto et al. / Journal of Power Sources 327 (2016) 327e332
performances (including rate capability, cyclability and polarization between discharge and charge voltage profiles) of the Na2FeP2O7// NaTi2(PO4)3 full-cell with 2 M Na2SO4 or 4 M NaClO4 were improved relative to those obtained with the non-aqueous electrolyte. This is attributed to the higher relative dielectric constant and lower viscosity of water than organic solvent. Conversely, the Na2FeP2O7//NaTi2(PO4)3 full-cell with 4 M NaNO3 exhibited a large irreversible capacity due to H2 gas evolution and corrosive side reactions. On the other hand, it may be difficult for the NaClO4 salts to use in market due to their explosiveness and oxidizing abilities. We therefore propose a Na2FeP2O7//NaTi2(PO4)3 full-cell in conjunction with 2 M Na2SO4 as a promising aqueous sodium-ion battery candidate. Acknowledgements This work was financially supported by Elements Strategy Initiative for Catalysts and Batteries Project, MEXT, Japan. References [1] P.G. Balakrishnan, R. Ramesh, T.P. Kumar, J. Power Sources 155 (2006) 401e414. [2] J. Zhao, L. Zhao, N. Dimov, S. Okada, T. Nishida, J. Electrochem. Soc. 160 (6) (2013) A3077eA3081. [3] A.D. Tevar, J.F. Whitacre, J. Electrochem. Soc. 157 (7) (2010) A870eA875. [4] J.F. Whitacre, T. Wiley, S. Shanbhag, Y. Wenzhuo, A. Mohamed, S.E. Chun, E. Weber, D. Blackwood, E. Lynch-Bell, J. Gulakowski, C. Smith, D. Humphreys, J. Power Sources 213 (2012) 255e264. [5] S.I. Park, I. Gocheva, S. Okada, J. Yamaki, J. Electrochem. Soc. 158 (10) (2011) A1067eA1070. [6] W. Wu, A. Mohamed, J.F. Whitacre, J. Electrochem. Soc. 160 (3) (2013)
A497eA504. [7] Z. Li, D.B. Ravnsbæk, K. Xiang, Y.M. Chiang, Electrochem. Commun. 44 (2014) 12e15. [8] H. Qin, Z.P. Song, H. Zhan, Y.H. Zhou, J. Power Sources 249 (2014) 367e372. [9] Z. Li, D. Young, K. Xiang, W.C. Carter, Y.M. Chiang, Adv. Energy Mater. 3 (2013) 290e294. [10] W. Wu, S. Shabhag, J. Chang, A. Rutt, J.F. Whitacre, J. Electrochem. Soc. 162 (6) (2015) A803eA808. [11] X. Wu, Y. Cao, X. Ai, J. Qian, H. Yang, Electrochem. Commun. 31 (2013) 145e148. [12] C.D. Wessells, S.V. Peddada, R.A. Huggins, Y. Cui, Nano Lett. 11 (2011) 5421e5425. [13] X.Y. Wu, M.Y. Sun, Y.F. Shen, J.F. Qian, Y.L. Cao, X.P. Ai, H.X. Yang, Chem. Sus. Chem. 7 (2014) 407e411. [14] W. Song, X. Ji, Y. Zhu, H. Zhu, F. Li, J. Chen, F. Lu, Y. Yao, C.E. Banks, Chem. Electro. Chem. 1 (2014) 871e876. [15] C. Deng, S. Zhang, Z. Dong, Y. Shang, Nano Energy 4 (2014) 49e55. [16] Y. Liu, B.H. Zhang, S.Y. Xiao, L.L. Liu, Z.B. Wen, Y.P. Wu, Electrochim. Acta 116 (2014) 512e517. [17] S.M. Oh, S.T. Myung, J. Hassoun, B. Scrosati, Y.K. Sun, Electrochem. Commun. 22 (2012) 149e152. [18] M. Vujkovi c, S. Mentus, J. Power Sources 247 (2014) 184e188. [19] P. Barpanda, G. Liu, C.D. Ling, M. Tamaru, M. Avdeev, S.C. Chung, Y. Yamada, A. Yamada, Chem. Mater 25 (2013) 3480e3487. [20] T. Honma, T. Togashi, N. Ito, T. Komatsu, J. Ceram. Soc. Jpn. 120 (8) (2012) 344e346. [21] H. Kim, R.A. Shakoor, C. Park, S.Y. Lim, J.S. Kim, Y.N. Jo, W. Cho, K. Miyasaka, R. Kahraman, Y. Jung, J.W. Choi, Adv. Funct. Mater. 23 (2013) 1147e1155. [22] Y.H. Jung, C.H. Lim, J.H. Kim, D.K. Kim, RSC Adv. 4 (2014) 9799e9802. [23] C.Y. Chen, K. Matsumoto, T. Nohira, R. Hagiwara, J. Electrochem. Soc. 162 (1) (2015) A176eA180. [24] C.R. Mariappan, C. Galven, M.-P. Crosnier-Lopez, F. Le Berre, O. Bohnke, J. Solid State Chem. 179 (2006) 450e456. [25] J.Y. Luo, W.J. Cui, P. He, Y.Y. Xia, Nat. Chem. 2 (2010) 760e765. [26] K. Kuratani, I. Kishimoto, Y. Nishida, R. Kondo, H.T. Takeshita, H. Senoh, T. Kiyobayashi, J. Electrochem. Soc. 163 (2016) H417eH425. [27] H. Wang, K. Huang, Y. Zeng, S. Yang, L. Chen, Electrochem. Acta 52 (2007) 3280e3285.