Letter pubs.acs.org/NanoLett
Advanced Na[Ni0.25Fe0.5Mn0.25]O2/C−Fe3O4 Sodium-Ion Batteries Using EMS Electrolyte for Energy Storage Seung-Min Oh,†,□ Seung-Taek Myung,‡,□ Chong Seung Yoon,§ Jun Lu,⊥ Jusef Hassoun,∥ Bruno Scrosati,*,∥ Khalil Amine,*,⊥,¶ and Yang-Kook Sun*,†,¶ †
Department of Energy Engineering, Hanyang University, Seoul 133-791, South Korea Department of Nano Engineering, Sejong University, Seoul 143-747, South Korea § Department of Materials Science and Engineering, Hanyang University, Seoul 133-791, South Korea ∥ Department of Chemistry, University of Rome Sapienza, 00185, Rome, Italy ⊥ Chemical Sciences and Engineering Division, Argonne National Laboratory, 9700 South Cass Avenue, Lemont, Illinois 60439, United States ¶ Chemistry Department, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia ‡
S Supporting Information *
ABSTRACT: While much research effort has been devoted to the development of advanced lithium-ion batteries for renewal energy storage applications, the sodium-ion battery is also of considerable interest because sodium is one of the most abundant elements in the Earth’s crust. In this work, we report a sodium-ion battery based on a carbon-coated Fe3O4 anode, Na[Ni0.25Fe0.5Mn0.25]O2 layered cathode, and NaClO4 in fluoroethylene carbonate and ethyl methanesulfonate electrolyte. This unique battery system combines an intercalation cathode and a conversion anode, resulting in high capacity, high rate capability, thermal stability, and much improved cycle life. This performance suggests that our sodium-ion system is potentially promising power sources for promoting the substantial use of low-cost energy storage systems in the near future. KEYWORDS: Carbon-coated iron oxide anode, layered nickel−iron-manganese cathode, sodium perchlorate, ethyl methanesulfonate, sodium-ion battery
L
ithium is an attractive battery material because it is the lightest metal, but its resources are concentrated in large quantities mainly in South America with an associated risk of dependence on a single geographic area.1 In addition, the price of lithium-ion batteries is affected by the high cost of their active materials, commonly based on cobalt compounds. As a result, there is an increasing interest worldwide in developing low cost and sustainable energy storage systems that do not use lithium or cobalt. Among them, rechargeable sodium batteries, due to the almost infinite supply of the metal, are the most appealing as immediate alternatives to lithium batteries. Recently reported results2−21 have triggered an increasing industrial and academic interest in sodium-ion batteries operating at room temperature. Pioneering work by Hagenmuller et al.2−6 and Yamamoto et al.7 introduced O3-type NaMO2 (M = Ni, Co, and Fe) in which Na+ could be inserted into a host structure. Since then, interest in developing sodium batteries declined because of the fast development and market domination of lithium-ion batteries. However, recent increasing demands for low-cost energy storage have renewed interest in sodium-based batteries and, accordingly, the O3 type NaMO2 © 2014 American Chemical Society
compounds are being revisited as a cathode material due to the facile synthesis and structural stability,8−16 including NaNi0.5Mn0.5O2,10 NaCoO2,11 NaCrO2,12 NaMnxM1−xO2 (M = Co, Ni),13−15 Na[Ni1/3Fe1/3O1/3]O2,16 and NaxVO2.17−19 At present, the electrochemical performance of these cathode materials has been limited due to the lack of electrolytes capable of withstanding voltages above 3.9 V versus Na/Na+.10 In response, we have developed a sodium-ion battery that has an electrolyte of NaClO4 in a mixture of ethyl methanesulfonate electrolyte (EMS) and fluoroethylene carbonate (FEC) additive for increasing the stability and conductivity, along with an anode of carbon-coated Fe3O4 and a cathode of O3-type layered Na[Ni0.25Fe0.5Mn0.25]O2. The electrolyte is a key component if sodium-ion batteries are to become competitive with or surpass lithium-ion batteries. The main requirement is a wide anodic stability window in order to allow complete Received: January 7, 2014 Revised: February 7, 2014 Published: February 13, 2014 1620
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
iron oxide, C−Fe3O4 conversion material. Graphite, the most common anode for lithium batteries, cannot be used in sodium batteries due to the large difference in ionic radii between lithium and sodium, which induces exfoliation rather than intercalation with dendritic growth of reactive Na metal. Hard carbon has been the most common alternative anode material, but its capacity is severely limited by the applied current density. A capacity of 320 mAh (g-carbon)−1 can be achieved at 0.1 C, while this capacity dramatically dropped to 130 mAh g−1 at 2 C-rate, which is mainly due to the slow kinetics of Na+ carrier transfer.26 In contrast, the carbon-coated Fe3O4 selected as the anode material in this work, which is obtained through a hydrothermal synthesis approach (see Supporting Information Experimental Section), has a theoretical capacity of about 920 mAh g−1 when used in a lithium cell and undergoes a conversion reaction (Fe3O4 + 8Li+ + 8e− → 4Li2O + 3Fe). Various conversion materials (CoOx, CuO, NiO, and FeO) for lithium batteries have been already introduced by Tarascon et al. in 2000.27 A similar conversion reaction is also expected to occur when this iron oxide is used as an anode in sodium cells, which will be proved in the next section of this paper. Although the feasibility of Fe3O4 as the cathode or anode material in Na ion cells with sodium intercalation or conversion chemistry have been explored recently,28,29 the various analysis to identify the phenomenon for the anode with a conversion chemistry for Na-ion battery is needed. Therefore, our approach involving the conversion anode and EMS-based electrolyte for Na-ion cells is unprecedented and provides further incentives for exploration of advanced materials for Na-ion batteries. In addition, the density of Fe3O4 (5.17g cc−1) is more than 2 times higher than amorphous carbon (1.8 g cc−1). As a result, the loading of Fe3O4 in the electrode is far higher than hard carbon. This translates to a significant increase in the energy density of Fe3O4 in a cell level compared to hard carbon. The scanning electron microscopy (SEM) image, shown in Figure 2a, reveals that the as-prepared C−Fe3O4 consists of agglomerates of primary particles having a size of about 50−70 nm. This morphology is confirmed by transmission electron microscopy (TEM) analysis, showing that the nanosized particles are uniformly distributed and coated by a carbon layer (Supporting Information Figure S3a). The Rietveld refinement of the X-ray diffraction (XRD) pattern indicates a crystalline Fe3O4 inverse spinel structure with Fd3m space group (see Figure 2b) with a calculated lattice parameter of 8.3923(1) Å, which is in good agreement with that reported in the literature.30 The electric conductivity was determined to be 2.3 × 10−3 S cm−1 at room temperature. The electrochemical properties of the C−Fe3O4 anode were examined by galvanostatical cycling of the cell in a sodium cell using the NaClO4/EMS/FEC electrolyte (Figure 2c). A capacity of 450 mAh (g-Fe3O4)−1 is obtained at the first discharge (reduction), and the voltage profile (dQ/dV curve, inset of Figure 2c) shows the following four distinct plateaus: above 1.3, 1.3, 1.3−0.75, and below 0.75 V. We interpret this profile as follows: (i) the capacity delivered at above 1.3 V is associated with the formation of a solid-electrolyte interface (SEI); (ii) the sharp peak with negligible capacity at 1.3 V is Na+ insertion into the Fe3O4 framework; (iii) the broad peak in the 1.3−0.75 V range could be attributed to decomposition of electrolyte; and (iv) the reduction of oxide to metal along with the related conversion reaction is dominant below 0.75 V. In this region, the observed capacity is about 250 mAh g−1. In the following charge (oxidation) process, only 200 mAh (g-
desodiation of the cathode. A common choice has been 1 M sodium perchlorate in propylene carbonate (NaClO4−PC) with selected additives.22−24 For instance, as suggested by Komaba et al.,25 the addition of a small amount of FEC to the PC-based solvent contributes to improved cycleability of sodium batteries, possibly because of a positive effect induced by the fluorination of the anode surfaces. Characteristics of Ethyl Methanesulfonate-Based Electrolyte. Figure 1a shows a sweep voltammogram of a
Figure 1. (a) Linear sweep voltammetry of a sodium cell using two electrolytes, namely, 1 M NaClO4 in PC + 2 vol % FEC and 1 M NaClO4 in EMS + 2 vol % FEC (blue circle denotes open-circuit potential) and (b) conductivity Arrhenius plot of the same electrolytes at various temperatures.
sodium cell using 1 M NaClO4 in PC + 2 vol % FEC electrolyte. No current flow is observed up to the voltage limit of the decomposition of PC-based electrolyte (4.5 V vs Na/ Na+). The anodic stability of the electrolyte can be further increased to 5.6 V versus Na/Na+ by replacing the PC with EMS, as shown in Figure 1a. Furthermore, the EMS slightly increases the ionic conductivity of the electrolyte; for example, it is 6.0 × 10−3 S cm−1 at 25 °C compared with 5.7 × 10−3 S cm−1 for the electrolyte with PC (Figure 1b). We confirmed the better suitability of EMS-containing electrolyte (dissolving 1 M LiPF6 and 1 M NaClO4) through the testing of 5 V Li-ion cells with LiNi0.45Fe0.1Mn1.45O4 and Na[Ni0.25Fe0.5Mn0.25]O2 electrodes (Supporting Information Figures S1 and S2). Therefore, we selected 1 M NaClO4 in EMS/FEC solution as the electrolyte for the sodium-ion battery developed in this work. Electrochemical and Physical Properties of Anode Materials. As the anode material, we selected carbon-modified 1621
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 2. (a) SEM image and (b) Rietveld refinement of XRD patterns for C−Fe3O4 nanosized powders (inset: structural image of Fe3O4 spinel structure, red sphere denotes oxygen). (c) Voltage profile of the first charge−discharge cycle of a Na/C−Fe3O4 cell at a constant current density of 20 mA g−1 (0.1 C-rate, 0−2 V) and related differential capacity curves (inset). (d) Ex situ XRD analysis and magnetic hysteresis loops (inset) of the cell at various stages of its first charge−discharge cycle shown in (c). (e) Scheme of the electrochemical process of the cell: Fe3O4 (left) is converted to Fe (right) metal on charge and vice versa on discharge and (f) its cycle performance and rate capability at various rates; for this test, the cell was first charged at 0.1 C-rate to 0 V and then discharged at rates varying from 0.1 to 10 C until 2.0 V at room temperature.
Fe3O4)−1 is recovered, indicating the occurrence of a large irreversible capacity. We may then describe the process of the first discharge as Fe3O4 + 1.74Na+ + 1.74e− → 0.78Fe3O4 + 0.65Fe + 0.87Na2O. The utilization of the active material is below 22%, likely due to an early termination of the conversion reaction, as evident from the finding that the conversion voltage in the sodium cell is lower than that observed in the parent lithium cell, that is, below ∼0.75−0 V versus ∼0.7−0.8 V.23 In addition, the resulting voltage curve for the Na system is sloppy whereas plateaus are observed in lithium system, indicating that the conversion reaction would be less favored in Na system than in Li system. To address this issue, we presodiated the C− Fe3O4 electrode by direct contacting with Na metal in the presence of the electrolyte. Under this condition, the electrode keeps the same capacity of 200 mAh g−1 both in charge and in discharge. As seen in Figure 2c, the C−Fe3O4 electrode
sodiated up to Na 1.74 Fe3 O4 after the first charge. In combination with the SQUID and XRD data, the sodiated amount of sodium is almost 0.44 mol per Fe3O4. The structure of the C−Fe3O4 electrode after presodiation is the same as the XRD result of “c” in Figure 2d. Figure 2d shows the ex situ XRD patterns of the sodiated Fe3O4 electrode upon charge and discharge. The intensities of the peaks changed, though not their position. This finding indicates that the electrochemical process is not of the intercalation type but rather of the expected conversion type, as confirmed by the fact that Na2O, that is, one of the products of the conversion reaction, is observed at the end of discharge. The as-fabricated Fe3O4 electrode is ferromagnetic, as evident from the black curve in the magnetic hysteresis loop displayed in the inset of Figure 2d. The curve remains basically unchanged after discharge to 0.75 V (red line c), indicating that 1622
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 3. (a) SEM image and (b) Rietveld refinement of XRD patterns of Na[Ni0.25Fe0.5Mn0.25]O2 powders. (c) Voltage profile of the first charge− discharge cycle of a Na/Na[Ni0.25Fe0.5Mn0.25]O2 cell at a constant current density of 13 mA g−1 (0.1 C-rate, 2.1−3.9 V) and related differential capacity curves (inset). (d) Ex situ XRD analysis of the cell at various stages of its first charge−discharge cycle shown in (c). (e) Scheme of the electrochemical process of the cell and (f) its cycle performance and rate capability at various C-rates and at various temperatures.
metal to Fe3O4 on charge, confirming the electrochemical process of the C−Fe3O4 electrode and its reversibility in a sodium cell (Figure 2e). Continuous cycling of the presodiated Fe3O4 electrodes for 50 cycles was carried out at three current rates: 20 mA g−1 (0.1 C-rate), 100 mA g−1 (0.5 C-rate), and 200 mA g−1 (1 C-rate). The results, reported in Figure 2f and Supporting Information Figure S10, show a very stable capacity delivery over 50 cycles at these rates. In tests to rates as high as 10 C (2000 mA g−1), the anode still exhibited a high capacity of 157 mAh g−1, which is much better compared to that of hard carbon.26 This behavior is in a way of a surprise, because a rapid drop of capacity is frequently observed for conversion-based metal oxide electrodes in Li-ion cells due to their repetitive volume expansion and contraction upon cycling.26 The cycled electrode did not show such a morphological change (Supporting Information Figure S3), assuring electrode stability. Though a high capacity of 350 mAh g−1 has recently
SEI formation and electrolyte decomposition are the major reactions in this voltage window. The magnetization at 5000 Oe, however, quickly rises for the fully discharged electrode (blue line d), confirming that the conversion of Fe3O4 to Fe metal does occur at potentials below 0.75 V. This conclusion is further confirmed by TEM analysis, which clearly shows the presence of Fe metal nanoparticles at the end of the discharge (see Supporting Information Figure S3b). Taking into account that the saturation magnetization for pure Fe is equal to 220 emu g−1 (also considering oxidation of the sample during handling for the measurement), we estimated that the overall fraction of Fe in the fully discharged electrode is 18%, consistent with calculated value of 22% utilization of the Fe3O4 electrode from the electrochemical test. Therefore, the large irreversible capacity is understandable. The XRD patterns of Figure 2d also reveal the signature of Na2O formation, as well as the progressive recovery of the Fe 1623
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
steps represented by a plateau to 3.2 V and a slope to 3.9 V. The differentiated capacity curve reflects those steps by showing two peaks at 2.93 and 3.13 V (inset to Figure 3c) representative of the plateau region, followed by a broad peak at higher voltages, associated with the slope region. Good reversibility of this electrochemical process is observed by the near overlap of the charge and discharge profiles (Figure 3c), which implies fast kinetics with small cycling polarization. Figure 3d shows an ex situ XRD analysis carried out in the course of the first cycle as a function of the Na content in Na1−δ[Ni0.25Fe0.5Mn0.25]O2. The patterns slightly broaden and weaken with increasing amount of transferred δ charge. In this fashion, the original O3 type layer structure is observed until δ = 0.2 upon charge; as the Na ion extraction proceeds to δ = 0.4, a new phase, having an interlayer distance longer than that of the pristine O3-type layer structure, appears and becomes dominant at the end of charge, δ = 0.62. The XRD pattern of the new phase can be indexed to a P3-type monoclinic structure with space group of C2/m. As a consequence of the stronger repulsion of oxygen in the NaO6 layers, the interlayer distance expands upon Na+ extraction in the NFM electrode owing to the lack of positive charged Na ions. Hence, it is reasonable to assume that the fast Na ion diffusion (measured as high as ∼10−9 S cm−1, Supporting Information Figure S6) is primarily due to the expanded interlayer spacing in this region, where the P3-type monoclinic structure is present as the major phase. The average oxidation state of transition metal elements such as Ni and Fe varies as the desodiation proceeds with Ni2+ converting to Ni4+ and Fe3+ to Fe4+ at the end of charge, while almost no change occurs for the oxidation state of Mn (Supporting Information Figure S7). On discharge, the P3-type monoclinic phase is transformed back into the original O3-type layer structure, and the average oxidation states of the transition metals are recovered to their original values in the XANES spectra (Supporting Information Figure S7). These results confirm the high reversibility of the NFM electrochemical process, as schematized in Figure 3e. The Na-NFM half cells show good capacity retention (Figure 3f, Supporting Information Figure S11): 90.4% at 25 °C and 86% at 55 °C after 50 cycles at the 1 C-rate. At the 10 C-rate, the electrode still delivers about 85 mAh g−1. Even at the low temperature of 0 °C, the capacity retention remained approximately 85 and 80% at the 0.5 C-rate and 1 C-rate, respectively. We attribute the high rate capability of our NFM electrode to the weak electrostatic interaction that, by causing longer alkali−metal−oxygen bonding, favors an easy Na+ diffusion across its structure. These are also affected by the better stability and conductivity of EMS-based electrolyte, which led to the cycle retention, rate capability, and Coulombic efficiency of NFM cell. Finally, the differential scanning calorimetry analysis presented in Supporting Information Figure S8 shows that the exothermic temperature of the deeply desodiated Na0.38[Ni0.25Fe0.5Mn0.25]O2 electrode appears fairly high, 280 °C with small heat generation (ca. 400 J g−1), likely favored by the presence of stable Mn4+ in the crystal framework. All these aspects demonstrate that NFM is a very promising cathode material to be coupled with an Fe3O4 anode in 1 M NaClO4 in EMS/FEC electrolyte to provide an advanced and novel sodium-ion battery. Through the above-reported results, we found that the individual component in the current Na-ion full cell, including anode (Fe3O4), and cathode (Na[Ni0.25Fe0.5Mn0.25]O2) materials have shown much improved electrochemical properties with
reported for an amorphous carbon with a small amount of nano Fe3O4 incorporated (52.7% of Fe3O4) The resulting capacity retention and Coulombic efficiency were very poor because these Fe3O4 nanoparticles were attached on amorphous carbon matrix of which the carbon structure is disintegrated due to the insertion of large Na+ ions into the amorphous carbon and continuous decomposition of electrolyte on discharge.29 The significantly improved capacity retention and rate capability of the C−Fe3O4 anode can be explained by the shallow conversion reaction, occurring with only a 22% efficiency, leaving a major fraction of unreacted Fe3O4, which, in turn, can act as a buffer to retain the structural stability of electrode from the volume expansion during cycling. In addition, other parameters contribute to assuring a high rate capability, such as (i) the nanosize of the electrode particles and their modification by the carbon coating and (ii) the high electric conductivity. Electrochemical and Physical Properties of Cathode Materials. As the cathode material, we selected the nickeliron-manganese sodium oxide Na[Ni0.25Fe0.5Mn0.25]O2 that calcined from synthesized [Ni0.25Fe0.5Mn0.25](OH)2 precursor via coprecipitation method, hereafter referred to NFM. This precursor and calcined materials exhibits spherical morphologies (Figure 3a, Supporting Information Figure S4) and a high tap density of 2.1 g cm−3. The as-synthesized NFM powder is stable in air, which is different from NaFeO2. The Rietveld refinement of the XRD data (Figure 3b) indicates that NFM displays a highly crystalline, O3 type, Na1.000[Ni0.249Fe0.502Mn0.249]O2 layer structure, as confirmed by chemical analysis. A similar Na[Ni1/3Fe1/3Mn1/3]O2 material was recently reported by Kim et al.;16 however, our material differs from that of reported by Kim et al. in terms of particle size, tap density, morphology, and phase homogeneity. In addition, the coprecipitation synthetic approach adopted in this study is a fairly facile synthesis, which is different from that reported by Kim. et al. (see Supporting Information Experimental Section). The calculated lattice parameters are a = 2.9907(1) Å and c = 15.9889(5) Å (Supporting Information Table S1), which are in between those for NaFeO2 (a = 3.03 Å and c = 16.11 Å)7 and Na[Ni0.5Mn0.5]O2 (a = 2.9901(1) Å and c = 15.9329 Å).10 On the basis of the above data, we may describe NFM as formed by a solid solution between Na[Ni0.5Mn0.5]O2 and NaFeO2, with a molar ratio of 1:1. From the increment of lattice parameters, compared with those of Na[Ni0.5Mn0.5]O2, we assume that the iron has an oxidation state of +3 with high spin state (ionic radius: 0.645 Å).32 This assumption is confirmed by the X-ray absorption nearest edge spectra (XANES, Supporting Information Figure S5), leading to the conclusion that the average oxidation states of Ni, Fe, and Mn are +2, +3, and +4, respectively. The roomtemperature electric conductivity of the NFM powder is approximately 4-fold greater than that of iron-free Na[Ni0.5Mn0.5]O2 material, namely, 6.6 × 10−7 S cm−1 versus 1.6 × 10−7 S cm−1. Obviously, the incorporation of iron into the crystal structure is effective in enhancing the electric conductivity, which is associated with a low band gap energy (about 2.5 eV for Fe2O3).31 Figure 3c shows the voltage profile for the first charge (sodiation)-discharge (desodiation) cycle from 2.1 to 3.9 V versus Na+/Na at a current density of 13 mA g−1 (0.1 C-rate), which demonstrates that the electrode delivers a capacity of approximately 140 mAh g−1 with an efficiency of 93%. The charge and discharge curves evolve with two reversible, distinct 1624
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 4. (a) Scheme of the C-Fe3O4/NaClO4-EMS + 2 vol% FEC/Na[Ni0.25Fe0.5Mn0.25]O2 full sodium-ion battery. (b) First two charge−discharge cycles at 0.1 C-rate (13 mA g−1, 0.5−3.6 V) at 25 °C. (c) Cycle performance for the Fe3O4/Na[Ni0.25Fe0.5Mn0.25]O2 full cell. (d) Rate capability at various rates.
achieved by setting the electrode density ratio of anode/ cathode to 1.2. The cell is then cathode limited, and its overall reaction is expected to be a combined intercalation (cathode) and conversion (anode) chemistry, as shown below (where x is the presodiated amount of Na)
high voltage electrolyte (EMS-based electrolyte). The full sodium-ion battery using each of the materials proposed in this work is sketched in Figure 4a. Prior to cell fabrication, a presodiation of the C−Fe3O4 anode was performed so as to reduce its irreversible capacity (Figure 2c). The cell balance was charge
Na[Ni 0.25Fe0.5Mn 0.25]O2 + NaxFe3O4 [ooooooooZ Na1 ‐ δ[Ni 0.25Fe0.5Mn 0.25]O2 + Naδ + xFe3O4 (Na 2O + Fe + Fe3O4 ) discharge
Figure 4b illustrates the voltage profiles of the full cell for the first charge and discharge curves under 1 C-rate. The difference in the first and the second cycle is ascribed to the different discharge profile derived from the presodiation (Figure 2c). The battery operates reversibly around 2.4 V, delivering a capacity of about 130 mAh (g-NFM)−1. Under these test conditions, the battery offers a good capacity retention for extensive cycling, namely, 82.8% at the 100th cycle and 76.1% at the 150th cycle with a Coulombic efficiency approaching 100% (see Figure 4c, average Coulombic efficiency: 99.614%). The battery was first charged to 3.6 V (constant current− constant voltage mode) with a current density of 13 mA g−1 (0.1 C-rate) and then discharged to 0.5 V at rates extending from 0.1 to 10 C. The rate capability varies from 130 mAh g−1 at 0.1 C, to 120 mAh g−1 at 1 C, to 72 mAh g−1 at 10 C (Supporting Information Figure S9). Such rate capability of the full cell is even comparable to the state-of-the-art lithium-ion battery. The good rate capability and cycling performances of the present Fe3O4/NFM Na-ion battery is mainly attributed to the high-rate available positive and negative active materials,
(1)
which are supported by their structural stability and good electric conductivity, as well as the stability of the EMS based electrolyte developed in this study. In summary, we reported a thermally stable NFM cathode material and improved electrochemical performance of Fe3O4/ NFM sodium-ion battery exploiting a unique combination of intercalation and conversion electrode chemistry in the presence of a novel electrolyte based on NaClO4 in EMS/ FEC. We are able to demonstrate that this battery pack offers very promising performance in terms of capacity and rate capability for a cycle life extending to over 150 cycles. Such performance is believed to associate with the structural stability and the high electric conductivity of the selected electrode materials, as well as the fast ion transport and wide electrochemical window of the EMS based electrolyte. In addition, the battery is based on sodium, an abundant, hence low cost, material. Overall, our sodium-ion system appears to be one of the potentially promising for low-cost energy storage in the near future. 1625
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
■
Letter
(15) Oh, S.-M.; Myung, S.-T.; Jang, M.-W.; Scrosati, B.; Hassoun, J.; Sun, Y.-K. Phys. Chem. Chem. Phys. 2013, 15, 3827−3833. (16) Kim, D.; Lee, E.; Slater, M.; Lu, W.; Rood, S.; Johnson, C. S. Electrochem. Commun. 2012, 18, 66−69. (17) Szajwaj, O.; Gaudin, E.; Weill, F.; Darriet, J.; Delmas, C. Inorg. Chem. 2009, 48, 9147−9154. (18) Queen, T. M.; Stephens, P. W.; Huang, Q.; Klimczuk, T.; Ronning, F.; Cava, R. J. Phy. Rev. Lett. 2008, 101, 166402(1)− 166402(4). (19) Hamani, D.; Ati, M.; Tarascon, J.-M.; Rozier, P. Electrochem. Commun. 2011, 13, 938−941. (20) Ponrouch, A.; Dedyvere, R.; Monti, D.; Demet, A. E.; Mba, J. M. A.; Croquennec, L.; Masquelier, C.; Johansson, P.; Palacin, M. R. Energy Environ. Sci. 2013, 6, 2361−2369. (21) Barker, J.; Saidi, M. Y.; Swoyer, J. L. Electrochem. Solid State Lett. 2003, 6, A1−A4. (22) Xia, X.; Dahn, J. R. J. Electrochem. Soc. 2012, 159, A515−A519. (23) Ponrouch, A.; Marchante, E.; Courty, M.; Tarascon, J.-M.; Palacin, M. R. Energy Environ. Sci. 2012, 5, 8572−8583. (24) Komaba, S.; Ishikawa, T.; Yabuuchi, N.; Murata, W.; Ito, A.; Ohsawa, Y. ACS Appl. Mater. Interfaces 2011, 3, 4165−4168. (25) Komaba, S.; Murata, W.; Ishikawa, T.; Yabuuchi, N.; Ozeki, T.; Nakayama, T.; Ogata, A.; Gotoh, K.; Fujiwara, K. Adv. Funct. Mater. 2011, 21, 3859−3867. (26) Ponrouch, A.; Goni, A. R.; Palacin, M. R. Electrochem. Commun. 2013, 27, 85−88. (27) Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.-M. Nature 2000, 407, 496−499. (28) Komaba, S.; Mikumo, T.; Yabuuchi, N.; Ogata, A.; Yoshida, H.; Yamada, Y. J. Electrochem. Soc. 2010, 157, A60−A65. (29) Hariharan, S.; Saravanan, K.; Ramar, V.; Balaya, P. Phys. Chem. Chem. Phys. 2013, 15, 2945−2953. (30) Kim, H. S.; Baek, S. H.; Jang, M.-W.; Sun, Y.-K.; Yoon, C. S. J. Electrochem. Soc. 2012, 159, A325−A329. (31) Noh, K. J.; Kim, B.-R.; Yoon, G.-J.; Jung, S.-C.; Kang, W.; Kim, S.-J. Electron. Mater. Lett. 2012, 8, 345−350. (32) Shannon, R. D. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1976, A32, 751−767. (33) Lee, M.-H.; Kang, Y.-J.; Myung, S.-T.; Sun, Y.-K. Electrochim. Acta 2004, 50, 939−948.
ASSOCIATED CONTENT
S Supporting Information *
Experimental procedure, additional electrochemical test, TEM, Rietveld refinement, XANES, GITT, and DSC results. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: (Y.S.) [email protected]. *E-mail: (B.S.) [email protected]. *E-mail: (K.A.) [email protected]. Author Contributions □
S.-M.O. and S.-T.M. contributed equally.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Human Resources Development program (No. 20124010203310) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Trade, Industry, and Energy and by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 2009-0092780). J.L. was supported by the Department of Energy (DOE) Office of Energy Efficiency and Renewable Energy (EERE) Postdoctoral Research Award under the EERE Vehicles Technology Program administered by the Oak Ridge Institute for Science and Education (ORISE) for the DOE. ORISE is managed by Oak Ridge Associated Universities (ORAU) under DOE contract number DE-AC05-06OR23100. Argonne National Laboratory is operated for the U.S. Department of Energy by UChicago Argonne, LLC, under contract DE-AC02-06CH11357.
■
REFERENCES
(1) Scrosati, B.; Garche, J. J. Power Sources 2010, 195, 2419−2430. (2) Fouassier, C.; Matejka, G.; Reau, J.-M.; Hagenmuller, P. J. Solid State Chem. 1973, 6, 532−537. (3) Delmas, C.; Fouassier, C.; Hagenmuller, P. Physica B+C 1980, 99, 81−85. (4) Braconnier, J.-J.; Delmas, C.; Fouassier, C.; Hagenmuller, P. Mater. Res. Bull. 1980, 15, 1797−1804. (5) Molenda, J.; Delmas, C.; Hagenmuller, P. Solid State Ionics 1983, 9−10, 431−435. (6) Maazaz, A.; Delmas, C.; Hagenmuller, P. J. Inclusion Phenom. 1983, 1, 45−51. (7) Takeda, Y.; Nakahara, K.; Nishijima, M.; Imanishi, N.; Yamamoto, O. Mater. Res. Bull. 1994, 29, 659−666. (8) Lu, Z.; Dahn, J. R. J. Electrochem. Soc. 2001, 148, A1225−A1229. (9) Yabuuchi, N.; Kajiyama, M.; Iwatate, J.; Nishikawa, H.; Hitomi, S.; Okuyama, R.; Usui, R.; Yamada, Y.; Komaba, S. Nat. Mater. 2012, 11, 512−517. (10) Komaba, S.; Nakayama, T.; Ogata, A.; Shimizu, T.; Takei, C.; Takada, S.; Hokura, A.; Nakai, I. ECS Trans. 2009, 16, 43−45. (11) Berthelot, R.; Carlier, D.; Delmas, C. Nat. Mater. 2011, 10, 74− 80. (12) Komaba, S.; Takei, C.; Nakayama, T.; Ogata, A.; Yabuuchi, N. Electrochem. Commun. 2010, 12, 355−358. (13) Saadoune, I.; Maazaz, A.; Menetrier, M.; Delmas, C. J. Solid State Chem. 1996, 122, 111−117. (14) Kim, D.; Kang, S.-H.; Slater, M.; Rood, S.; Vaughey, J. T.; Karan, N.; Balasubramanian, M.; Johnson, C. S. Adv. Energy Mater. 2011, 1, 333−336. 1626
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Advanced Na[Ni0.25Fe0.5Mn0.25]O2/C−Fe3O4 Sodium-Ion Batteries Using EMS Electrolyte for Energy Storage Seung-Min Oh,†,□ Seung-Taek Myung,‡,□ Chong Seung Yoon,§ Jun Lu,⊥ Jusef Hassoun,∥ Bruno Scrosati,*,∥ Khalil Amine,*,⊥,¶ and Yang-Kook Sun*,†,¶ †
Department of Energy Engineering, Hanyang University, Seoul 133-791, South Korea Department of Nano Engineering, Sejong University, Seoul 143-747, South Korea § Department of Materials Science and Engineering, Hanyang University, Seoul 133-791, South Korea ∥ Department of Chemistry, University of Rome Sapienza, 00185, Rome, Italy ⊥ Chemical Sciences and Engineering Division, Argonne National Laboratory, 9700 South Cass Avenue, Lemont, Illinois 60439, United States ¶ Chemistry Department, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia ‡
S Supporting Information *
ABSTRACT: While much research effort has been devoted to the development of advanced lithium-ion batteries for renewal energy storage applications, the sodium-ion battery is also of considerable interest because sodium is one of the most abundant elements in the Earth’s crust. In this work, we report a sodium-ion battery based on a carbon-coated Fe3O4 anode, Na[Ni0.25Fe0.5Mn0.25]O2 layered cathode, and NaClO4 in fluoroethylene carbonate and ethyl methanesulfonate electrolyte. This unique battery system combines an intercalation cathode and a conversion anode, resulting in high capacity, high rate capability, thermal stability, and much improved cycle life. This performance suggests that our sodium-ion system is potentially promising power sources for promoting the substantial use of low-cost energy storage systems in the near future. KEYWORDS: Carbon-coated iron oxide anode, layered nickel−iron-manganese cathode, sodium perchlorate, ethyl methanesulfonate, sodium-ion battery
L
ithium is an attractive battery material because it is the lightest metal, but its resources are concentrated in large quantities mainly in South America with an associated risk of dependence on a single geographic area.1 In addition, the price of lithium-ion batteries is affected by the high cost of their active materials, commonly based on cobalt compounds. As a result, there is an increasing interest worldwide in developing low cost and sustainable energy storage systems that do not use lithium or cobalt. Among them, rechargeable sodium batteries, due to the almost infinite supply of the metal, are the most appealing as immediate alternatives to lithium batteries. Recently reported results2−21 have triggered an increasing industrial and academic interest in sodium-ion batteries operating at room temperature. Pioneering work by Hagenmuller et al.2−6 and Yamamoto et al.7 introduced O3-type NaMO2 (M = Ni, Co, and Fe) in which Na+ could be inserted into a host structure. Since then, interest in developing sodium batteries declined because of the fast development and market domination of lithium-ion batteries. However, recent increasing demands for low-cost energy storage have renewed interest in sodium-based batteries and, accordingly, the O3 type NaMO2 © 2014 American Chemical Society
compounds are being revisited as a cathode material due to the facile synthesis and structural stability,8−16 including NaNi0.5Mn0.5O2,10 NaCoO2,11 NaCrO2,12 NaMnxM1−xO2 (M = Co, Ni),13−15 Na[Ni1/3Fe1/3O1/3]O2,16 and NaxVO2.17−19 At present, the electrochemical performance of these cathode materials has been limited due to the lack of electrolytes capable of withstanding voltages above 3.9 V versus Na/Na+.10 In response, we have developed a sodium-ion battery that has an electrolyte of NaClO4 in a mixture of ethyl methanesulfonate electrolyte (EMS) and fluoroethylene carbonate (FEC) additive for increasing the stability and conductivity, along with an anode of carbon-coated Fe3O4 and a cathode of O3-type layered Na[Ni0.25Fe0.5Mn0.25]O2. The electrolyte is a key component if sodium-ion batteries are to become competitive with or surpass lithium-ion batteries. The main requirement is a wide anodic stability window in order to allow complete Received: January 7, 2014 Revised: February 7, 2014 Published: February 13, 2014 1620
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
iron oxide, C−Fe3O4 conversion material. Graphite, the most common anode for lithium batteries, cannot be used in sodium batteries due to the large difference in ionic radii between lithium and sodium, which induces exfoliation rather than intercalation with dendritic growth of reactive Na metal. Hard carbon has been the most common alternative anode material, but its capacity is severely limited by the applied current density. A capacity of 320 mAh (g-carbon)−1 can be achieved at 0.1 C, while this capacity dramatically dropped to 130 mAh g−1 at 2 C-rate, which is mainly due to the slow kinetics of Na+ carrier transfer.26 In contrast, the carbon-coated Fe3O4 selected as the anode material in this work, which is obtained through a hydrothermal synthesis approach (see Supporting Information Experimental Section), has a theoretical capacity of about 920 mAh g−1 when used in a lithium cell and undergoes a conversion reaction (Fe3O4 + 8Li+ + 8e− → 4Li2O + 3Fe). Various conversion materials (CoOx, CuO, NiO, and FeO) for lithium batteries have been already introduced by Tarascon et al. in 2000.27 A similar conversion reaction is also expected to occur when this iron oxide is used as an anode in sodium cells, which will be proved in the next section of this paper. Although the feasibility of Fe3O4 as the cathode or anode material in Na ion cells with sodium intercalation or conversion chemistry have been explored recently,28,29 the various analysis to identify the phenomenon for the anode with a conversion chemistry for Na-ion battery is needed. Therefore, our approach involving the conversion anode and EMS-based electrolyte for Na-ion cells is unprecedented and provides further incentives for exploration of advanced materials for Na-ion batteries. In addition, the density of Fe3O4 (5.17g cc−1) is more than 2 times higher than amorphous carbon (1.8 g cc−1). As a result, the loading of Fe3O4 in the electrode is far higher than hard carbon. This translates to a significant increase in the energy density of Fe3O4 in a cell level compared to hard carbon. The scanning electron microscopy (SEM) image, shown in Figure 2a, reveals that the as-prepared C−Fe3O4 consists of agglomerates of primary particles having a size of about 50−70 nm. This morphology is confirmed by transmission electron microscopy (TEM) analysis, showing that the nanosized particles are uniformly distributed and coated by a carbon layer (Supporting Information Figure S3a). The Rietveld refinement of the X-ray diffraction (XRD) pattern indicates a crystalline Fe3O4 inverse spinel structure with Fd3m space group (see Figure 2b) with a calculated lattice parameter of 8.3923(1) Å, which is in good agreement with that reported in the literature.30 The electric conductivity was determined to be 2.3 × 10−3 S cm−1 at room temperature. The electrochemical properties of the C−Fe3O4 anode were examined by galvanostatical cycling of the cell in a sodium cell using the NaClO4/EMS/FEC electrolyte (Figure 2c). A capacity of 450 mAh (g-Fe3O4)−1 is obtained at the first discharge (reduction), and the voltage profile (dQ/dV curve, inset of Figure 2c) shows the following four distinct plateaus: above 1.3, 1.3, 1.3−0.75, and below 0.75 V. We interpret this profile as follows: (i) the capacity delivered at above 1.3 V is associated with the formation of a solid-electrolyte interface (SEI); (ii) the sharp peak with negligible capacity at 1.3 V is Na+ insertion into the Fe3O4 framework; (iii) the broad peak in the 1.3−0.75 V range could be attributed to decomposition of electrolyte; and (iv) the reduction of oxide to metal along with the related conversion reaction is dominant below 0.75 V. In this region, the observed capacity is about 250 mAh g−1. In the following charge (oxidation) process, only 200 mAh (g-
desodiation of the cathode. A common choice has been 1 M sodium perchlorate in propylene carbonate (NaClO4−PC) with selected additives.22−24 For instance, as suggested by Komaba et al.,25 the addition of a small amount of FEC to the PC-based solvent contributes to improved cycleability of sodium batteries, possibly because of a positive effect induced by the fluorination of the anode surfaces. Characteristics of Ethyl Methanesulfonate-Based Electrolyte. Figure 1a shows a sweep voltammogram of a
Figure 1. (a) Linear sweep voltammetry of a sodium cell using two electrolytes, namely, 1 M NaClO4 in PC + 2 vol % FEC and 1 M NaClO4 in EMS + 2 vol % FEC (blue circle denotes open-circuit potential) and (b) conductivity Arrhenius plot of the same electrolytes at various temperatures.
sodium cell using 1 M NaClO4 in PC + 2 vol % FEC electrolyte. No current flow is observed up to the voltage limit of the decomposition of PC-based electrolyte (4.5 V vs Na/ Na+). The anodic stability of the electrolyte can be further increased to 5.6 V versus Na/Na+ by replacing the PC with EMS, as shown in Figure 1a. Furthermore, the EMS slightly increases the ionic conductivity of the electrolyte; for example, it is 6.0 × 10−3 S cm−1 at 25 °C compared with 5.7 × 10−3 S cm−1 for the electrolyte with PC (Figure 1b). We confirmed the better suitability of EMS-containing electrolyte (dissolving 1 M LiPF6 and 1 M NaClO4) through the testing of 5 V Li-ion cells with LiNi0.45Fe0.1Mn1.45O4 and Na[Ni0.25Fe0.5Mn0.25]O2 electrodes (Supporting Information Figures S1 and S2). Therefore, we selected 1 M NaClO4 in EMS/FEC solution as the electrolyte for the sodium-ion battery developed in this work. Electrochemical and Physical Properties of Anode Materials. As the anode material, we selected carbon-modified 1621
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 2. (a) SEM image and (b) Rietveld refinement of XRD patterns for C−Fe3O4 nanosized powders (inset: structural image of Fe3O4 spinel structure, red sphere denotes oxygen). (c) Voltage profile of the first charge−discharge cycle of a Na/C−Fe3O4 cell at a constant current density of 20 mA g−1 (0.1 C-rate, 0−2 V) and related differential capacity curves (inset). (d) Ex situ XRD analysis and magnetic hysteresis loops (inset) of the cell at various stages of its first charge−discharge cycle shown in (c). (e) Scheme of the electrochemical process of the cell: Fe3O4 (left) is converted to Fe (right) metal on charge and vice versa on discharge and (f) its cycle performance and rate capability at various rates; for this test, the cell was first charged at 0.1 C-rate to 0 V and then discharged at rates varying from 0.1 to 10 C until 2.0 V at room temperature.
Fe3O4)−1 is recovered, indicating the occurrence of a large irreversible capacity. We may then describe the process of the first discharge as Fe3O4 + 1.74Na+ + 1.74e− → 0.78Fe3O4 + 0.65Fe + 0.87Na2O. The utilization of the active material is below 22%, likely due to an early termination of the conversion reaction, as evident from the finding that the conversion voltage in the sodium cell is lower than that observed in the parent lithium cell, that is, below ∼0.75−0 V versus ∼0.7−0.8 V.23 In addition, the resulting voltage curve for the Na system is sloppy whereas plateaus are observed in lithium system, indicating that the conversion reaction would be less favored in Na system than in Li system. To address this issue, we presodiated the C− Fe3O4 electrode by direct contacting with Na metal in the presence of the electrolyte. Under this condition, the electrode keeps the same capacity of 200 mAh g−1 both in charge and in discharge. As seen in Figure 2c, the C−Fe3O4 electrode
sodiated up to Na 1.74 Fe3 O4 after the first charge. In combination with the SQUID and XRD data, the sodiated amount of sodium is almost 0.44 mol per Fe3O4. The structure of the C−Fe3O4 electrode after presodiation is the same as the XRD result of “c” in Figure 2d. Figure 2d shows the ex situ XRD patterns of the sodiated Fe3O4 electrode upon charge and discharge. The intensities of the peaks changed, though not their position. This finding indicates that the electrochemical process is not of the intercalation type but rather of the expected conversion type, as confirmed by the fact that Na2O, that is, one of the products of the conversion reaction, is observed at the end of discharge. The as-fabricated Fe3O4 electrode is ferromagnetic, as evident from the black curve in the magnetic hysteresis loop displayed in the inset of Figure 2d. The curve remains basically unchanged after discharge to 0.75 V (red line c), indicating that 1622
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 3. (a) SEM image and (b) Rietveld refinement of XRD patterns of Na[Ni0.25Fe0.5Mn0.25]O2 powders. (c) Voltage profile of the first charge− discharge cycle of a Na/Na[Ni0.25Fe0.5Mn0.25]O2 cell at a constant current density of 13 mA g−1 (0.1 C-rate, 2.1−3.9 V) and related differential capacity curves (inset). (d) Ex situ XRD analysis of the cell at various stages of its first charge−discharge cycle shown in (c). (e) Scheme of the electrochemical process of the cell and (f) its cycle performance and rate capability at various C-rates and at various temperatures.
metal to Fe3O4 on charge, confirming the electrochemical process of the C−Fe3O4 electrode and its reversibility in a sodium cell (Figure 2e). Continuous cycling of the presodiated Fe3O4 electrodes for 50 cycles was carried out at three current rates: 20 mA g−1 (0.1 C-rate), 100 mA g−1 (0.5 C-rate), and 200 mA g−1 (1 C-rate). The results, reported in Figure 2f and Supporting Information Figure S10, show a very stable capacity delivery over 50 cycles at these rates. In tests to rates as high as 10 C (2000 mA g−1), the anode still exhibited a high capacity of 157 mAh g−1, which is much better compared to that of hard carbon.26 This behavior is in a way of a surprise, because a rapid drop of capacity is frequently observed for conversion-based metal oxide electrodes in Li-ion cells due to their repetitive volume expansion and contraction upon cycling.26 The cycled electrode did not show such a morphological change (Supporting Information Figure S3), assuring electrode stability. Though a high capacity of 350 mAh g−1 has recently
SEI formation and electrolyte decomposition are the major reactions in this voltage window. The magnetization at 5000 Oe, however, quickly rises for the fully discharged electrode (blue line d), confirming that the conversion of Fe3O4 to Fe metal does occur at potentials below 0.75 V. This conclusion is further confirmed by TEM analysis, which clearly shows the presence of Fe metal nanoparticles at the end of the discharge (see Supporting Information Figure S3b). Taking into account that the saturation magnetization for pure Fe is equal to 220 emu g−1 (also considering oxidation of the sample during handling for the measurement), we estimated that the overall fraction of Fe in the fully discharged electrode is 18%, consistent with calculated value of 22% utilization of the Fe3O4 electrode from the electrochemical test. Therefore, the large irreversible capacity is understandable. The XRD patterns of Figure 2d also reveal the signature of Na2O formation, as well as the progressive recovery of the Fe 1623
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
steps represented by a plateau to 3.2 V and a slope to 3.9 V. The differentiated capacity curve reflects those steps by showing two peaks at 2.93 and 3.13 V (inset to Figure 3c) representative of the plateau region, followed by a broad peak at higher voltages, associated with the slope region. Good reversibility of this electrochemical process is observed by the near overlap of the charge and discharge profiles (Figure 3c), which implies fast kinetics with small cycling polarization. Figure 3d shows an ex situ XRD analysis carried out in the course of the first cycle as a function of the Na content in Na1−δ[Ni0.25Fe0.5Mn0.25]O2. The patterns slightly broaden and weaken with increasing amount of transferred δ charge. In this fashion, the original O3 type layer structure is observed until δ = 0.2 upon charge; as the Na ion extraction proceeds to δ = 0.4, a new phase, having an interlayer distance longer than that of the pristine O3-type layer structure, appears and becomes dominant at the end of charge, δ = 0.62. The XRD pattern of the new phase can be indexed to a P3-type monoclinic structure with space group of C2/m. As a consequence of the stronger repulsion of oxygen in the NaO6 layers, the interlayer distance expands upon Na+ extraction in the NFM electrode owing to the lack of positive charged Na ions. Hence, it is reasonable to assume that the fast Na ion diffusion (measured as high as ∼10−9 S cm−1, Supporting Information Figure S6) is primarily due to the expanded interlayer spacing in this region, where the P3-type monoclinic structure is present as the major phase. The average oxidation state of transition metal elements such as Ni and Fe varies as the desodiation proceeds with Ni2+ converting to Ni4+ and Fe3+ to Fe4+ at the end of charge, while almost no change occurs for the oxidation state of Mn (Supporting Information Figure S7). On discharge, the P3-type monoclinic phase is transformed back into the original O3-type layer structure, and the average oxidation states of the transition metals are recovered to their original values in the XANES spectra (Supporting Information Figure S7). These results confirm the high reversibility of the NFM electrochemical process, as schematized in Figure 3e. The Na-NFM half cells show good capacity retention (Figure 3f, Supporting Information Figure S11): 90.4% at 25 °C and 86% at 55 °C after 50 cycles at the 1 C-rate. At the 10 C-rate, the electrode still delivers about 85 mAh g−1. Even at the low temperature of 0 °C, the capacity retention remained approximately 85 and 80% at the 0.5 C-rate and 1 C-rate, respectively. We attribute the high rate capability of our NFM electrode to the weak electrostatic interaction that, by causing longer alkali−metal−oxygen bonding, favors an easy Na+ diffusion across its structure. These are also affected by the better stability and conductivity of EMS-based electrolyte, which led to the cycle retention, rate capability, and Coulombic efficiency of NFM cell. Finally, the differential scanning calorimetry analysis presented in Supporting Information Figure S8 shows that the exothermic temperature of the deeply desodiated Na0.38[Ni0.25Fe0.5Mn0.25]O2 electrode appears fairly high, 280 °C with small heat generation (ca. 400 J g−1), likely favored by the presence of stable Mn4+ in the crystal framework. All these aspects demonstrate that NFM is a very promising cathode material to be coupled with an Fe3O4 anode in 1 M NaClO4 in EMS/FEC electrolyte to provide an advanced and novel sodium-ion battery. Through the above-reported results, we found that the individual component in the current Na-ion full cell, including anode (Fe3O4), and cathode (Na[Ni0.25Fe0.5Mn0.25]O2) materials have shown much improved electrochemical properties with
reported for an amorphous carbon with a small amount of nano Fe3O4 incorporated (52.7% of Fe3O4) The resulting capacity retention and Coulombic efficiency were very poor because these Fe3O4 nanoparticles were attached on amorphous carbon matrix of which the carbon structure is disintegrated due to the insertion of large Na+ ions into the amorphous carbon and continuous decomposition of electrolyte on discharge.29 The significantly improved capacity retention and rate capability of the C−Fe3O4 anode can be explained by the shallow conversion reaction, occurring with only a 22% efficiency, leaving a major fraction of unreacted Fe3O4, which, in turn, can act as a buffer to retain the structural stability of electrode from the volume expansion during cycling. In addition, other parameters contribute to assuring a high rate capability, such as (i) the nanosize of the electrode particles and their modification by the carbon coating and (ii) the high electric conductivity. Electrochemical and Physical Properties of Cathode Materials. As the cathode material, we selected the nickeliron-manganese sodium oxide Na[Ni0.25Fe0.5Mn0.25]O2 that calcined from synthesized [Ni0.25Fe0.5Mn0.25](OH)2 precursor via coprecipitation method, hereafter referred to NFM. This precursor and calcined materials exhibits spherical morphologies (Figure 3a, Supporting Information Figure S4) and a high tap density of 2.1 g cm−3. The as-synthesized NFM powder is stable in air, which is different from NaFeO2. The Rietveld refinement of the XRD data (Figure 3b) indicates that NFM displays a highly crystalline, O3 type, Na1.000[Ni0.249Fe0.502Mn0.249]O2 layer structure, as confirmed by chemical analysis. A similar Na[Ni1/3Fe1/3Mn1/3]O2 material was recently reported by Kim et al.;16 however, our material differs from that of reported by Kim et al. in terms of particle size, tap density, morphology, and phase homogeneity. In addition, the coprecipitation synthetic approach adopted in this study is a fairly facile synthesis, which is different from that reported by Kim. et al. (see Supporting Information Experimental Section). The calculated lattice parameters are a = 2.9907(1) Å and c = 15.9889(5) Å (Supporting Information Table S1), which are in between those for NaFeO2 (a = 3.03 Å and c = 16.11 Å)7 and Na[Ni0.5Mn0.5]O2 (a = 2.9901(1) Å and c = 15.9329 Å).10 On the basis of the above data, we may describe NFM as formed by a solid solution between Na[Ni0.5Mn0.5]O2 and NaFeO2, with a molar ratio of 1:1. From the increment of lattice parameters, compared with those of Na[Ni0.5Mn0.5]O2, we assume that the iron has an oxidation state of +3 with high spin state (ionic radius: 0.645 Å).32 This assumption is confirmed by the X-ray absorption nearest edge spectra (XANES, Supporting Information Figure S5), leading to the conclusion that the average oxidation states of Ni, Fe, and Mn are +2, +3, and +4, respectively. The roomtemperature electric conductivity of the NFM powder is approximately 4-fold greater than that of iron-free Na[Ni0.5Mn0.5]O2 material, namely, 6.6 × 10−7 S cm−1 versus 1.6 × 10−7 S cm−1. Obviously, the incorporation of iron into the crystal structure is effective in enhancing the electric conductivity, which is associated with a low band gap energy (about 2.5 eV for Fe2O3).31 Figure 3c shows the voltage profile for the first charge (sodiation)-discharge (desodiation) cycle from 2.1 to 3.9 V versus Na+/Na at a current density of 13 mA g−1 (0.1 C-rate), which demonstrates that the electrode delivers a capacity of approximately 140 mAh g−1 with an efficiency of 93%. The charge and discharge curves evolve with two reversible, distinct 1624
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
Letter
Figure 4. (a) Scheme of the C-Fe3O4/NaClO4-EMS + 2 vol% FEC/Na[Ni0.25Fe0.5Mn0.25]O2 full sodium-ion battery. (b) First two charge−discharge cycles at 0.1 C-rate (13 mA g−1, 0.5−3.6 V) at 25 °C. (c) Cycle performance for the Fe3O4/Na[Ni0.25Fe0.5Mn0.25]O2 full cell. (d) Rate capability at various rates.
achieved by setting the electrode density ratio of anode/ cathode to 1.2. The cell is then cathode limited, and its overall reaction is expected to be a combined intercalation (cathode) and conversion (anode) chemistry, as shown below (where x is the presodiated amount of Na)
high voltage electrolyte (EMS-based electrolyte). The full sodium-ion battery using each of the materials proposed in this work is sketched in Figure 4a. Prior to cell fabrication, a presodiation of the C−Fe3O4 anode was performed so as to reduce its irreversible capacity (Figure 2c). The cell balance was charge
Na[Ni 0.25Fe0.5Mn 0.25]O2 + NaxFe3O4 [ooooooooZ Na1 ‐ δ[Ni 0.25Fe0.5Mn 0.25]O2 + Naδ + xFe3O4 (Na 2O + Fe + Fe3O4 ) discharge
Figure 4b illustrates the voltage profiles of the full cell for the first charge and discharge curves under 1 C-rate. The difference in the first and the second cycle is ascribed to the different discharge profile derived from the presodiation (Figure 2c). The battery operates reversibly around 2.4 V, delivering a capacity of about 130 mAh (g-NFM)−1. Under these test conditions, the battery offers a good capacity retention for extensive cycling, namely, 82.8% at the 100th cycle and 76.1% at the 150th cycle with a Coulombic efficiency approaching 100% (see Figure 4c, average Coulombic efficiency: 99.614%). The battery was first charged to 3.6 V (constant current− constant voltage mode) with a current density of 13 mA g−1 (0.1 C-rate) and then discharged to 0.5 V at rates extending from 0.1 to 10 C. The rate capability varies from 130 mAh g−1 at 0.1 C, to 120 mAh g−1 at 1 C, to 72 mAh g−1 at 10 C (Supporting Information Figure S9). Such rate capability of the full cell is even comparable to the state-of-the-art lithium-ion battery. The good rate capability and cycling performances of the present Fe3O4/NFM Na-ion battery is mainly attributed to the high-rate available positive and negative active materials,
(1)
which are supported by their structural stability and good electric conductivity, as well as the stability of the EMS based electrolyte developed in this study. In summary, we reported a thermally stable NFM cathode material and improved electrochemical performance of Fe3O4/ NFM sodium-ion battery exploiting a unique combination of intercalation and conversion electrode chemistry in the presence of a novel electrolyte based on NaClO4 in EMS/ FEC. We are able to demonstrate that this battery pack offers very promising performance in terms of capacity and rate capability for a cycle life extending to over 150 cycles. Such performance is believed to associate with the structural stability and the high electric conductivity of the selected electrode materials, as well as the fast ion transport and wide electrochemical window of the EMS based electrolyte. In addition, the battery is based on sodium, an abundant, hence low cost, material. Overall, our sodium-ion system appears to be one of the potentially promising for low-cost energy storage in the near future. 1625
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
Nano Letters
■
Letter
(15) Oh, S.-M.; Myung, S.-T.; Jang, M.-W.; Scrosati, B.; Hassoun, J.; Sun, Y.-K. Phys. Chem. Chem. Phys. 2013, 15, 3827−3833. (16) Kim, D.; Lee, E.; Slater, M.; Lu, W.; Rood, S.; Johnson, C. S. Electrochem. Commun. 2012, 18, 66−69. (17) Szajwaj, O.; Gaudin, E.; Weill, F.; Darriet, J.; Delmas, C. Inorg. Chem. 2009, 48, 9147−9154. (18) Queen, T. M.; Stephens, P. W.; Huang, Q.; Klimczuk, T.; Ronning, F.; Cava, R. J. Phy. Rev. Lett. 2008, 101, 166402(1)− 166402(4). (19) Hamani, D.; Ati, M.; Tarascon, J.-M.; Rozier, P. Electrochem. Commun. 2011, 13, 938−941. (20) Ponrouch, A.; Dedyvere, R.; Monti, D.; Demet, A. E.; Mba, J. M. A.; Croquennec, L.; Masquelier, C.; Johansson, P.; Palacin, M. R. Energy Environ. Sci. 2013, 6, 2361−2369. (21) Barker, J.; Saidi, M. Y.; Swoyer, J. L. Electrochem. Solid State Lett. 2003, 6, A1−A4. (22) Xia, X.; Dahn, J. R. J. Electrochem. Soc. 2012, 159, A515−A519. (23) Ponrouch, A.; Marchante, E.; Courty, M.; Tarascon, J.-M.; Palacin, M. R. Energy Environ. Sci. 2012, 5, 8572−8583. (24) Komaba, S.; Ishikawa, T.; Yabuuchi, N.; Murata, W.; Ito, A.; Ohsawa, Y. ACS Appl. Mater. Interfaces 2011, 3, 4165−4168. (25) Komaba, S.; Murata, W.; Ishikawa, T.; Yabuuchi, N.; Ozeki, T.; Nakayama, T.; Ogata, A.; Gotoh, K.; Fujiwara, K. Adv. Funct. Mater. 2011, 21, 3859−3867. (26) Ponrouch, A.; Goni, A. R.; Palacin, M. R. Electrochem. Commun. 2013, 27, 85−88. (27) Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.-M. Nature 2000, 407, 496−499. (28) Komaba, S.; Mikumo, T.; Yabuuchi, N.; Ogata, A.; Yoshida, H.; Yamada, Y. J. Electrochem. Soc. 2010, 157, A60−A65. (29) Hariharan, S.; Saravanan, K.; Ramar, V.; Balaya, P. Phys. Chem. Chem. Phys. 2013, 15, 2945−2953. (30) Kim, H. S.; Baek, S. H.; Jang, M.-W.; Sun, Y.-K.; Yoon, C. S. J. Electrochem. Soc. 2012, 159, A325−A329. (31) Noh, K. J.; Kim, B.-R.; Yoon, G.-J.; Jung, S.-C.; Kang, W.; Kim, S.-J. Electron. Mater. Lett. 2012, 8, 345−350. (32) Shannon, R. D. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1976, A32, 751−767. (33) Lee, M.-H.; Kang, Y.-J.; Myung, S.-T.; Sun, Y.-K. Electrochim. Acta 2004, 50, 939−948.
ASSOCIATED CONTENT
S Supporting Information *
Experimental procedure, additional electrochemical test, TEM, Rietveld refinement, XANES, GITT, and DSC results. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: (Y.S.) [email protected]. *E-mail: (B.S.) [email protected]. *E-mail: (K.A.) [email protected]. Author Contributions □
S.-M.O. and S.-T.M. contributed equally.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Human Resources Development program (No. 20124010203310) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Trade, Industry, and Energy and by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 2009-0092780). J.L. was supported by the Department of Energy (DOE) Office of Energy Efficiency and Renewable Energy (EERE) Postdoctoral Research Award under the EERE Vehicles Technology Program administered by the Oak Ridge Institute for Science and Education (ORISE) for the DOE. ORISE is managed by Oak Ridge Associated Universities (ORAU) under DOE contract number DE-AC05-06OR23100. Argonne National Laboratory is operated for the U.S. Department of Energy by UChicago Argonne, LLC, under contract DE-AC02-06CH11357.
■
REFERENCES
(1) Scrosati, B.; Garche, J. J. Power Sources 2010, 195, 2419−2430. (2) Fouassier, C.; Matejka, G.; Reau, J.-M.; Hagenmuller, P. J. Solid State Chem. 1973, 6, 532−537. (3) Delmas, C.; Fouassier, C.; Hagenmuller, P. Physica B+C 1980, 99, 81−85. (4) Braconnier, J.-J.; Delmas, C.; Fouassier, C.; Hagenmuller, P. Mater. Res. Bull. 1980, 15, 1797−1804. (5) Molenda, J.; Delmas, C.; Hagenmuller, P. Solid State Ionics 1983, 9−10, 431−435. (6) Maazaz, A.; Delmas, C.; Hagenmuller, P. J. Inclusion Phenom. 1983, 1, 45−51. (7) Takeda, Y.; Nakahara, K.; Nishijima, M.; Imanishi, N.; Yamamoto, O. Mater. Res. Bull. 1994, 29, 659−666. (8) Lu, Z.; Dahn, J. R. J. Electrochem. Soc. 2001, 148, A1225−A1229. (9) Yabuuchi, N.; Kajiyama, M.; Iwatate, J.; Nishikawa, H.; Hitomi, S.; Okuyama, R.; Usui, R.; Yamada, Y.; Komaba, S. Nat. Mater. 2012, 11, 512−517. (10) Komaba, S.; Nakayama, T.; Ogata, A.; Shimizu, T.; Takei, C.; Takada, S.; Hokura, A.; Nakai, I. ECS Trans. 2009, 16, 43−45. (11) Berthelot, R.; Carlier, D.; Delmas, C. Nat. Mater. 2011, 10, 74− 80. (12) Komaba, S.; Takei, C.; Nakayama, T.; Ogata, A.; Yabuuchi, N. Electrochem. Commun. 2010, 12, 355−358. (13) Saadoune, I.; Maazaz, A.; Menetrier, M.; Delmas, C. J. Solid State Chem. 1996, 122, 111−117. (14) Kim, D.; Kang, S.-H.; Slater, M.; Rood, S.; Vaughey, J. T.; Karan, N.; Balasubramanian, M.; Johnson, C. S. Adv. Energy Mater. 2011, 1, 333−336. 1626
dx.doi.org/10.1021/nl500077v | Nano Lett. 2014, 14, 1620−1626
