- Email: [email protected]
Amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) composites as a cathode material for lithium ion batteries
Solid State Ionics 326 (2018) 48–51
Contents lists available at ScienceDirect
Solid State Ionics journal homepage: www.elsevier.com/locate/ssi
Amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) composites as a cathode material for lithium ion batteries ⁎
Ayuko Kitajoua, , Yuji Ishadob, Atsushi Inoishic, Shigeto Okadad,
T
⁎
a
Organization for Research Initiatives, Yamaguchi University, 2-16-1 Tokiwadai, Ube 755-8611, Japan Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan c Department of Applied Chemistry, Faculty of Engineering, Kyushu University, Motoka 744, Nishi-ku, Fukuoka 819-0395, Japan d Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Amorphous cathode material Dry ball-milling method Fluorosulfate compound
Although the synthesis is not so easy, LiFeSO4F, a high-voltage iron-based cathode, is attractive by virtue of its low cost and small environmental impact. We prepared amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) using the dry ballmilling method. Moreover, we succeeded to synthesize tavorite-type LiFeSO4F by sintering amorphous LiF-FeSO4 at 280 °C. Amorphous 1.3LiF-FeSO4 exhibited the highest reversible capacity, of about 130 mAh g−1, among all the xLiF-FeSO4 series (1 ≤ x ≤ 2), with an average voltage of 3.5 V. In addition, we found that the obtained amorphous xLiF-FeSO4 cathodes have excellent cyclability and rate capability, also.
1. Introduction Beyond Li-ion batteries are being developed to power electric vehicles and to efficiently utilize renewable energies, such as solar and wind power. In particular, researchers worldwide are intensely pursuing novel electrode materials for further advances in energy density. As potential cathodes for beyond Li-ion batteries, iron-based cathode materials have attracted attention by virtue of their low cost and the abundance of iron resources. Among these materials, LiFePO4 has a high operating voltage of 3.3 V corresponding to Fe2+/Fe3+ redox by the inductive effect of PO43− polyanions [1]. Lithium iron sulfate Li2Fe (SO4)2 [2], which substitutes sulfate SO42− with phosphate PO43− anions, shows the highest operating voltage of 3.75 V (vs. Li+/Li) among iron-based cathode materials, although its theoretical capacity is restricted by the high molecular weight of SO4. On the other hand, the theoretical capacity can be increased to 150 mAh g−1 from 102 mAh g−1 by changing from Li2Fe(SO4)2 to LiFeSO4F. This LiFeSO4F has been reported to have two types of crystal structures, tavorite-type and triplite-type LiFeSO4F, which showed operating voltages of 3.6 V and 3.75 V, respectively [3,4]. However, the reported LiMSO4F (M = Fe, Co, Ni) is not stable at the high temperatures (over 500 °C) used in the normal solid-state method. In addition, the other reported LiFeSO4F also required a special synthesis route such as ionic liquid, vacuum condition, or FeSO4 hydrate [5–7]. Because these synthesis routes are very costly, it is necessary to develop a simple process for synthesis of LiFeSO4F. Triplite-type LiFeSO4F has already been obtained ⁎
from nonhydrate FeSO4 and LiF by the dry ball-milling method [8]. However, its rechargeable capacity was 100 mAh g−1, corresponding to < 0.7 electron reaction per mole, and the calculated energy density was only 350 Wh kg−1. In this work, in order to develop a simple and low-cost synthesis method of LiFeSO4F and to improve the cathode properties, we used the dry ball-milling method to prepare amorphous LiF-FeSO4 having the same chemical composition as LiFeSO4F. Moreover, it is already known that the electric conductivity for solid electrolyte can improve by increasing Li concentration in amorphous compound. Therefore, we are interested in determining whether this theory is effective for amorphous cathode materials. So, we also prepared amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) with the dry ball-milling method and evaluated its cathode properties against Li metal anodes in Li-salt electrolytes. 2. Experiment Amorphous xLiF-FeSO4 composites (x = 1.0, 1.2, 1.3, 1.5, 1.7 and 2.0) were prepared by dry ball-milling. Mixtures of LiF (Wako Pure Chemical Industries) and FeSO4 with a molar ratio of xLiF:FeSO4 were put in an Ar-filled atmosphere control container with ϕ3-ZrO2 balls (ca. 40 g). The mixtures were ball-milled using a planetary mill (Fritsch, Pulverisette7) at 600 rpm under ambient Ar for 6 h. Here, to obtain FeSO4 as a starting material, FeSO4·7H2O (Wako Pure Chemical Industries) was sintered at 300 °C for 12 h under Ar. To obtain a uniform xLiF-FeSO4 (1 ≤ x ≤ 2) and carbon composite, the obtained
Corresponding authors. E-mail addresses: [email protected] (A. Kitajou), [email protected] (S. Okada).
https://doi.org/10.1016/j.ssi.2018.09.007 Received 8 June 2018; Received in revised form 7 September 2018; Accepted 13 September 2018 0167-2738/ © 2018 Elsevier B.V. All rights reserved.
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 1. (a) The XRD profiles of the obtained amorphous xLiF-FeSO4 (1 ≤ x ≤ 2), (b) EDS mapping for the obtained LiF-FeSO4 sample.
Fig. 2. (a) Structure of triplite-LiFeSO4F and tavorite-LiFeSO4F, (b) DSC profile for the obtained amorphous LiF-FeSO4, (c) XRD profiles of the samples sintered at 280 °C and 350 °C.
morphology and EDS (Energy Dispersive X-ray Spectrometer) mapping were observed by using a transmission electron microscope (TEM; JEOL JEM 2100F). The cathode properties of the amorphous xLiF-FeSO4/C and the crystalline LiFeSO4F were evaluated with a 2032 coin-type cell using 1 M LiPF6/EC:DMC = 1:1 in volume (Tomiyama Pure Chemical Industries) and a polypropylene separator (3501, Celgard) against lithium metal (Honjo Metal). The cathode pellets to evaluate the electrochemical properties were fabricated by mixing the xLiF-FeSO4/C composite powder with a 5 wt% polytetrafluoroethylene (PTFE) Teflon binder (Polyflon PTFE F-104; Daikin Industries, Ltd.) and punched into disks (ca. 30 mg weight and 10 mm diameter).
products were subjected to the carbon composite process twice. The obtained amorphous xLiF-FeSO4 was ball-milled with 5 wt% acetylene black (AB, Denki Kagaku) at 600 rpm for 6 h. The product was ballmilled again with 20 wt% AB in Ar. Crystalline LiFeSO4F was obtained from the amorphous LiF-FeSO4/C. To determine the sintering temperature, we measured the temperature profiles of amorphous LiFFeSO4/C using the Thermo Plus TG-DSC 8230L system (Rigaku). To obtain crystalline LiFeSO4F, amorphous xLiF-FeSO4 (x = 1.0) was sintered at 280 °C or 350 °C in a sealed SUS container. The obtained powders were characterized using powder X-ray diffraction (XRD, 50 kV and 300 mA, Cu Kα, Rigaku TTRIII). The particle size, particle 49
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 3. (a) Initial and second charge/discharge curves of amorphous xLiF-FeSO4 cathodes (x = 1, 1.3, 1.5, and 2) at a rate of 0.2 mA cm−2 between 2.5 V and 4.3 V, (b) the relationship between discharge capacity (or the number of reacted electrons) and stoichiometry x in xLiF-FeSO4.
FeSO4 was obtained with the dry ball-milling method. However, LiFeSO4F had two types of crystal structure, a tavorite-type and a triplite-type, as shown in Fig. 2 (a). In addition, according to a previous report [9], tavorite-type LiFeSO4F cannot be obtained from LiF and nonhydrate FeSO4. Therefore, we tried to sinter this amorphous sample. First, we measured the DSC profile of the obtained amorphous LiFFeSO4 to estimate the suitable sintering temperatures. From the DSC profile (Fig. 2 (b)), the obtained amorphous xLiF-FeSO4 (x = 1.0) showed a crystallization temperature of 265 °C and an exothermic peak at 300 °C. On the basis of these results, the obtained amorphous xLiFFeSO4 (x = 1.0) was sintered at 280 °C and 350 °C. Fig. 2 (c) compares the XRD profiles of the samples sintered at 280 °C and 350 °C. The sample sintered at 280 °C was indexed in the triclinic system with space group P-1 (tavorite-type LiFeSO4F). This result demonstrated that the heating of amorphous LiF-FeSO4 at 280 °C gave tavorite-type LiFeSO4F, successfully. On the other hand, the sample sintered at 350 °C was indexed in the monoclinic system with space group P21/c (triplite-type LiFeSO4F). This suggested that the observed exothermic peak at 300 °C is the phase transition from the tavorite phase to the triplite phase. Moreover, we tried to calculate the crystallite size of the obtained crystalline samples by using Scherrer equation from the XRD profiles. The crystallite sizes of the sample were 38 Å (tavorite-type LiFeSO4F) and 262 Å (triplite-type LiFeSO4F), respectively. The crystallite size of triplite-type LiFeSO4F was larger than that of tavorite-type LiFeSO4F, because triplite-type LiFeSO4F required a higher synthesis temperature. Fig. 3 (a) compares the initial and second charge-discharge curves for the obtained xLiF-FeSO4 cathode with x = 1.0, 1.3, 1.5, and 2.0 at a rate of 0.2 mA cm−2. The electrochemical measurements were performed between 2.5 and 4.3 V at 25 °C. The obtained samples have a relatively small overpotential of 0.2 V, and the average voltage is 3.5 V. In a previous report, triplite-LiFeSO4F prepared by the dry ball-milling method had an overpotential of about 0.5 V [8]. These results suggest that the cathode properties of the obtained samples are improved by the preparation of a uniform active material and the carbon to form a composite. Moreover, the initial charge and discharge capacities of xLiF-FeSO4 (x = 1.3) were 120 mAh g−1 and 131 mAh g−1, respectively. Fig. 3 (b) shows the relationship between initial discharge capacity and stoichiometry x in xLiF-FeSO4 (solid line), and that between the number of reacted electrons and stoichiometry x (broken line). The
Fig. 4. (a) Initial and second charge/discharge curves of the obtained crystalline samples from the amorphous xLiF-FeSO4 (x = 1) at a rate of 0.2 mA cm−2 between 2.5 V and 4.3 V. (b) The calculated dQ/dE curves from charge/discharge profile of the obtained amorphous and crystalline samples.
3. Results and discussion The obtained xLiF-FeSO4/C was characterized by XRD measurement, as shown in Fig. 1 (a). Since the diffraction peak of the obtained sample was a halo peak over a high background, it was considered to be consisted of amorphous or nano-particles. In addition, the diffraction peak of LiF remained as an unreacted substance in the LiF rich composition (x > 1.5). It suggests that some excess LiF cannot react with FeSO4. Fig. 1 (b) shows a TEM image of the obtained amorphous LiFFeSO4/C as well as element mapping by STEM-EDS. The obtained sample had a primary particle size of ca. 100–200 nm, and all elements were dispersed uniformly in all particles. Namely, the amorphous LiF50
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 5. (a) The cyclability and (b) rate capability for the obtained amorphous and crystalline samples between 2.5 V and 4.3 V.
0.1 mA cm−2 to 1.0 mA cm−2. Here, these measurements were carried out by using different cells at various current densities. The initial discharge capacities at a rate of 1.0 mA cm−2 were 97 mAh g−1 (x = 1.0), 125 mAh g−1 (x = 1.3), 122 mAh g−1 (tavorite type), and 112 mAh g−1 (triplite type), respectively. Although this measurement was performed with a pellet-type electrode having a thickness of over 100 μm, the discharge capacities barely deteriorated even at rate of 1.0 mA cm−2. In particular, the capacity maintenance factor (Q (1.0 mA cm−2)/Q (0.1 mA cm−2)) was 93% in x = 1.3. Here, Q is the initial discharge capacity. Though xLiF-FeSO4 (x = 1.3) was an amorphous sample, this cathode exhibited better rate capability than the crystalline samples.
results suggested that the initial discharge capacity increased in x < 1.3, and this tendency was the opposite that in x > 1.3. When x was larger than 1.3, the rechargeable capacity was decreased with increasing x in xLiF-FeSO4 at x > 1.3. The calculated theoretical capacity of xLiF-FeSO4 was decreasing from 150 mAh g−1 (x = 1.0) to 132 mAh g−1 (x = 2.0) with increasing x in xLiF-FeSO4. In addition, the Fe2+/Fe3+ redox of all of the included iron in amorphous xLiF-FeSO4 contributed to this charge-discharge reaction. These results suggested that the addition of LiF more than x = 1.3 contributed to the reduction of the rechargeable capacity of amorphous xLiF-FeSO4 rather than the improvement of cathode properties. In the case of amorphous solid electrolyte, Li conductivity is improved by high Li concentration in unit volume. On the other hand, Li diffusion can readily occur in Li-excess disordered materials, since the Li diffusion energy in such materials, which have a locally Li-rich environment, is decreased as the result of the much weaker electrostatic repulsion [10]. In our previous report [11], we have also elucidated that this theory can be effective against not only Li-excess disordered materials but also inverse spinel materials. Therefore, we expected that the reversible capacity of LiF-excess amorphous samples was higher than that of xLiF-FeSO4 (x = 1.0), which has the same chemical composition as LiFeSO4F. Fig. 4 (a) shows the initial and second charge-discharge curves for tavorite-type and triplite-type LiFeSO4F. The initial charge-discharge capacities were 122/132 mAh g−1 (tavorite type) and 111/ 116 mAh g−1 (triplite type), and the average voltages of the crystalline LiFeSO4F were 3.6 V (tavorite type) and 3.7 V (triplite type). These results were in good agreement with reported values in literatures [3,4]. High operating voltage of these cathode materials occurred by inductive effect of SO4 anion. In particular, the higher operating voltage of triplite-LiFeSO4 may be caused by Li/Fe disordering at octahedral site and the local structure of iron which are octahedrally coordinated by four oxygen and two fluorine. In addition, the overpotential of both crystalline samples was quite small as shown in Fig. 4 (b). On the other hand, the triplite LiFeSO4F prepared by dry ball-milling method has overpotential of about 0.5 V [8], but in our case, it is only 0.2 V. Therefore, we believe that the Li diffusion in the obtained amorphous xLiF-FeSO4 was improved by a locally Li-rich environment. Fig. 5 (a) summarizes the cyclability of the obtained amorphous and crystalline samples at a rate of 0.2 mA cm−2 between 2.5 and 4.3 V. The capacity retentions of the xLiF-FeSO4, tavorite-type, and triplite-type cathodes were 101% (x = 1.0), 100% (x = 1.3), 101% (tavorite type), and 100% (triplite type) even after 30 cycles, respectively. All samples showed excellent cyclability. The rate capabilities of the amorphous xLiF-FeSO4 (x = 1.0, 1.3), tavorite type, and triplite type between 2.5 and 4.3 V (Fig. 5(b)) were investigated with various currents ranging from
4. Conclusions Mechanical co-milling of LiF and nonhydrate FeSO4 successfully gave amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) systems. In addition, we elucidated that tavorite and triplite-type LiFeSO4F can be separately obtained by heating amorphous LiF-FeSO4 at 280 °C and 350 °C, respectively. The amorphous xLiF-FeSO4 cathode has a small overpotential and an average voltage of 3.5 V. Moreover, the xLiF-FeSO4 (x = 1.3) cathode had the highest rechargeable capacity among the amorphous xLiF-FeSO4 series (1 < x < 2), as well as excellent cyclability and rate capability. Although the generally amorphous cathode did not have good electrochemical properties, the Li-excess amorphous cathode can overcome this drawback by percolating the Li network in a locally Lirich environment. References [1] A.K. Padhi, K.S. Nanjundaswamy, C. Masquelier, S. Okada, J.B. Goodenough, J. Electrochem. Soc. 144 (1997) 1609. [2] L. Lander, M. Reynaud, G. Rousse, M.T. Sougrati, C.L. Robert, R.J. Messinger, M. Deschamps, J.-M. Tarascon, Chem. Mater. 26 (2014) 4178. [3] N. Recham, J.-N. Chotard, L. Dupont, C. Delacourt, W. Walker, M. Armand, J.M. Tarascon, Nat. Mater. 9 (2010) 68. [4] P. Barpanda, M. Ati, B.C. Melot, G. Rousse, J.-N. Chotard, M.-L. Doublet, M.T. Sougrati, S.A. Corr, J.-C. Jumas, J.-M. Tarascon, Nat. Mater. 10 (2011) 772. [5] L. Liu, B. Shang, X.-J. Huang, Prog. Nat. Sci. 21 (2011) 211. [6] L. Zhang, J.-M. Tarascon, M.T. Sougrati, G. Rousse, G. Chen, J. Mater. Chem. A 3 (2015) 16988. [7] P. Barpanda, N. Recham, J.-N. Chotard, K. Djellab, W. Walker, M. Armand, J.M. Tarascon, J. Mater. Chem. A 20 (2010) 1659. [8] M. Ati, M. Sathiya, S. Boulineau, M. Reynaud, A. Abakumov, G. Rousse, B. Melot, G.V. Tendeloo, J.-M. Tarascon, J. Am. Chem. Soc. 134 (2012) 18380. [9] M. Kim, Y. Jung, B. Kang, J. Mater. Chem. A 3 (2015) 7583. [10] J. Lee, A. Urban, X. Li, D. Su, G. Hautier, G. Ceder, Science 343 (2014) 519. [11] A. Kitajou, J. Yoshida, S. Nakanishi, Y. Matsuda, R. Kanno, T. Okajima, S. Okada, J. Power Sources 302 (2016) 204.
51
Contents lists available at ScienceDirect
Solid State Ionics journal homepage: www.elsevier.com/locate/ssi
Amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) composites as a cathode material for lithium ion batteries ⁎
Ayuko Kitajoua, , Yuji Ishadob, Atsushi Inoishic, Shigeto Okadad,
T
⁎
a
Organization for Research Initiatives, Yamaguchi University, 2-16-1 Tokiwadai, Ube 755-8611, Japan Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan c Department of Applied Chemistry, Faculty of Engineering, Kyushu University, Motoka 744, Nishi-ku, Fukuoka 819-0395, Japan d Institute for Materials Chemistry and Engineering, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Amorphous cathode material Dry ball-milling method Fluorosulfate compound
Although the synthesis is not so easy, LiFeSO4F, a high-voltage iron-based cathode, is attractive by virtue of its low cost and small environmental impact. We prepared amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) using the dry ballmilling method. Moreover, we succeeded to synthesize tavorite-type LiFeSO4F by sintering amorphous LiF-FeSO4 at 280 °C. Amorphous 1.3LiF-FeSO4 exhibited the highest reversible capacity, of about 130 mAh g−1, among all the xLiF-FeSO4 series (1 ≤ x ≤ 2), with an average voltage of 3.5 V. In addition, we found that the obtained amorphous xLiF-FeSO4 cathodes have excellent cyclability and rate capability, also.
1. Introduction Beyond Li-ion batteries are being developed to power electric vehicles and to efficiently utilize renewable energies, such as solar and wind power. In particular, researchers worldwide are intensely pursuing novel electrode materials for further advances in energy density. As potential cathodes for beyond Li-ion batteries, iron-based cathode materials have attracted attention by virtue of their low cost and the abundance of iron resources. Among these materials, LiFePO4 has a high operating voltage of 3.3 V corresponding to Fe2+/Fe3+ redox by the inductive effect of PO43− polyanions [1]. Lithium iron sulfate Li2Fe (SO4)2 [2], which substitutes sulfate SO42− with phosphate PO43− anions, shows the highest operating voltage of 3.75 V (vs. Li+/Li) among iron-based cathode materials, although its theoretical capacity is restricted by the high molecular weight of SO4. On the other hand, the theoretical capacity can be increased to 150 mAh g−1 from 102 mAh g−1 by changing from Li2Fe(SO4)2 to LiFeSO4F. This LiFeSO4F has been reported to have two types of crystal structures, tavorite-type and triplite-type LiFeSO4F, which showed operating voltages of 3.6 V and 3.75 V, respectively [3,4]. However, the reported LiMSO4F (M = Fe, Co, Ni) is not stable at the high temperatures (over 500 °C) used in the normal solid-state method. In addition, the other reported LiFeSO4F also required a special synthesis route such as ionic liquid, vacuum condition, or FeSO4 hydrate [5–7]. Because these synthesis routes are very costly, it is necessary to develop a simple process for synthesis of LiFeSO4F. Triplite-type LiFeSO4F has already been obtained ⁎
from nonhydrate FeSO4 and LiF by the dry ball-milling method [8]. However, its rechargeable capacity was 100 mAh g−1, corresponding to < 0.7 electron reaction per mole, and the calculated energy density was only 350 Wh kg−1. In this work, in order to develop a simple and low-cost synthesis method of LiFeSO4F and to improve the cathode properties, we used the dry ball-milling method to prepare amorphous LiF-FeSO4 having the same chemical composition as LiFeSO4F. Moreover, it is already known that the electric conductivity for solid electrolyte can improve by increasing Li concentration in amorphous compound. Therefore, we are interested in determining whether this theory is effective for amorphous cathode materials. So, we also prepared amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) with the dry ball-milling method and evaluated its cathode properties against Li metal anodes in Li-salt electrolytes. 2. Experiment Amorphous xLiF-FeSO4 composites (x = 1.0, 1.2, 1.3, 1.5, 1.7 and 2.0) were prepared by dry ball-milling. Mixtures of LiF (Wako Pure Chemical Industries) and FeSO4 with a molar ratio of xLiF:FeSO4 were put in an Ar-filled atmosphere control container with ϕ3-ZrO2 balls (ca. 40 g). The mixtures were ball-milled using a planetary mill (Fritsch, Pulverisette7) at 600 rpm under ambient Ar for 6 h. Here, to obtain FeSO4 as a starting material, FeSO4·7H2O (Wako Pure Chemical Industries) was sintered at 300 °C for 12 h under Ar. To obtain a uniform xLiF-FeSO4 (1 ≤ x ≤ 2) and carbon composite, the obtained
Corresponding authors. E-mail addresses: [email protected] (A. Kitajou), [email protected] (S. Okada).
https://doi.org/10.1016/j.ssi.2018.09.007 Received 8 June 2018; Received in revised form 7 September 2018; Accepted 13 September 2018 0167-2738/ © 2018 Elsevier B.V. All rights reserved.
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 1. (a) The XRD profiles of the obtained amorphous xLiF-FeSO4 (1 ≤ x ≤ 2), (b) EDS mapping for the obtained LiF-FeSO4 sample.
Fig. 2. (a) Structure of triplite-LiFeSO4F and tavorite-LiFeSO4F, (b) DSC profile for the obtained amorphous LiF-FeSO4, (c) XRD profiles of the samples sintered at 280 °C and 350 °C.
morphology and EDS (Energy Dispersive X-ray Spectrometer) mapping were observed by using a transmission electron microscope (TEM; JEOL JEM 2100F). The cathode properties of the amorphous xLiF-FeSO4/C and the crystalline LiFeSO4F were evaluated with a 2032 coin-type cell using 1 M LiPF6/EC:DMC = 1:1 in volume (Tomiyama Pure Chemical Industries) and a polypropylene separator (3501, Celgard) against lithium metal (Honjo Metal). The cathode pellets to evaluate the electrochemical properties were fabricated by mixing the xLiF-FeSO4/C composite powder with a 5 wt% polytetrafluoroethylene (PTFE) Teflon binder (Polyflon PTFE F-104; Daikin Industries, Ltd.) and punched into disks (ca. 30 mg weight and 10 mm diameter).
products were subjected to the carbon composite process twice. The obtained amorphous xLiF-FeSO4 was ball-milled with 5 wt% acetylene black (AB, Denki Kagaku) at 600 rpm for 6 h. The product was ballmilled again with 20 wt% AB in Ar. Crystalline LiFeSO4F was obtained from the amorphous LiF-FeSO4/C. To determine the sintering temperature, we measured the temperature profiles of amorphous LiFFeSO4/C using the Thermo Plus TG-DSC 8230L system (Rigaku). To obtain crystalline LiFeSO4F, amorphous xLiF-FeSO4 (x = 1.0) was sintered at 280 °C or 350 °C in a sealed SUS container. The obtained powders were characterized using powder X-ray diffraction (XRD, 50 kV and 300 mA, Cu Kα, Rigaku TTRIII). The particle size, particle 49
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 3. (a) Initial and second charge/discharge curves of amorphous xLiF-FeSO4 cathodes (x = 1, 1.3, 1.5, and 2) at a rate of 0.2 mA cm−2 between 2.5 V and 4.3 V, (b) the relationship between discharge capacity (or the number of reacted electrons) and stoichiometry x in xLiF-FeSO4.
FeSO4 was obtained with the dry ball-milling method. However, LiFeSO4F had two types of crystal structure, a tavorite-type and a triplite-type, as shown in Fig. 2 (a). In addition, according to a previous report [9], tavorite-type LiFeSO4F cannot be obtained from LiF and nonhydrate FeSO4. Therefore, we tried to sinter this amorphous sample. First, we measured the DSC profile of the obtained amorphous LiFFeSO4 to estimate the suitable sintering temperatures. From the DSC profile (Fig. 2 (b)), the obtained amorphous xLiF-FeSO4 (x = 1.0) showed a crystallization temperature of 265 °C and an exothermic peak at 300 °C. On the basis of these results, the obtained amorphous xLiFFeSO4 (x = 1.0) was sintered at 280 °C and 350 °C. Fig. 2 (c) compares the XRD profiles of the samples sintered at 280 °C and 350 °C. The sample sintered at 280 °C was indexed in the triclinic system with space group P-1 (tavorite-type LiFeSO4F). This result demonstrated that the heating of amorphous LiF-FeSO4 at 280 °C gave tavorite-type LiFeSO4F, successfully. On the other hand, the sample sintered at 350 °C was indexed in the monoclinic system with space group P21/c (triplite-type LiFeSO4F). This suggested that the observed exothermic peak at 300 °C is the phase transition from the tavorite phase to the triplite phase. Moreover, we tried to calculate the crystallite size of the obtained crystalline samples by using Scherrer equation from the XRD profiles. The crystallite sizes of the sample were 38 Å (tavorite-type LiFeSO4F) and 262 Å (triplite-type LiFeSO4F), respectively. The crystallite size of triplite-type LiFeSO4F was larger than that of tavorite-type LiFeSO4F, because triplite-type LiFeSO4F required a higher synthesis temperature. Fig. 3 (a) compares the initial and second charge-discharge curves for the obtained xLiF-FeSO4 cathode with x = 1.0, 1.3, 1.5, and 2.0 at a rate of 0.2 mA cm−2. The electrochemical measurements were performed between 2.5 and 4.3 V at 25 °C. The obtained samples have a relatively small overpotential of 0.2 V, and the average voltage is 3.5 V. In a previous report, triplite-LiFeSO4F prepared by the dry ball-milling method had an overpotential of about 0.5 V [8]. These results suggest that the cathode properties of the obtained samples are improved by the preparation of a uniform active material and the carbon to form a composite. Moreover, the initial charge and discharge capacities of xLiF-FeSO4 (x = 1.3) were 120 mAh g−1 and 131 mAh g−1, respectively. Fig. 3 (b) shows the relationship between initial discharge capacity and stoichiometry x in xLiF-FeSO4 (solid line), and that between the number of reacted electrons and stoichiometry x (broken line). The
Fig. 4. (a) Initial and second charge/discharge curves of the obtained crystalline samples from the amorphous xLiF-FeSO4 (x = 1) at a rate of 0.2 mA cm−2 between 2.5 V and 4.3 V. (b) The calculated dQ/dE curves from charge/discharge profile of the obtained amorphous and crystalline samples.
3. Results and discussion The obtained xLiF-FeSO4/C was characterized by XRD measurement, as shown in Fig. 1 (a). Since the diffraction peak of the obtained sample was a halo peak over a high background, it was considered to be consisted of amorphous or nano-particles. In addition, the diffraction peak of LiF remained as an unreacted substance in the LiF rich composition (x > 1.5). It suggests that some excess LiF cannot react with FeSO4. Fig. 1 (b) shows a TEM image of the obtained amorphous LiFFeSO4/C as well as element mapping by STEM-EDS. The obtained sample had a primary particle size of ca. 100–200 nm, and all elements were dispersed uniformly in all particles. Namely, the amorphous LiF50
Solid State Ionics 326 (2018) 48–51
A. Kitajou et al.
Fig. 5. (a) The cyclability and (b) rate capability for the obtained amorphous and crystalline samples between 2.5 V and 4.3 V.
0.1 mA cm−2 to 1.0 mA cm−2. Here, these measurements were carried out by using different cells at various current densities. The initial discharge capacities at a rate of 1.0 mA cm−2 were 97 mAh g−1 (x = 1.0), 125 mAh g−1 (x = 1.3), 122 mAh g−1 (tavorite type), and 112 mAh g−1 (triplite type), respectively. Although this measurement was performed with a pellet-type electrode having a thickness of over 100 μm, the discharge capacities barely deteriorated even at rate of 1.0 mA cm−2. In particular, the capacity maintenance factor (Q (1.0 mA cm−2)/Q (0.1 mA cm−2)) was 93% in x = 1.3. Here, Q is the initial discharge capacity. Though xLiF-FeSO4 (x = 1.3) was an amorphous sample, this cathode exhibited better rate capability than the crystalline samples.
results suggested that the initial discharge capacity increased in x < 1.3, and this tendency was the opposite that in x > 1.3. When x was larger than 1.3, the rechargeable capacity was decreased with increasing x in xLiF-FeSO4 at x > 1.3. The calculated theoretical capacity of xLiF-FeSO4 was decreasing from 150 mAh g−1 (x = 1.0) to 132 mAh g−1 (x = 2.0) with increasing x in xLiF-FeSO4. In addition, the Fe2+/Fe3+ redox of all of the included iron in amorphous xLiF-FeSO4 contributed to this charge-discharge reaction. These results suggested that the addition of LiF more than x = 1.3 contributed to the reduction of the rechargeable capacity of amorphous xLiF-FeSO4 rather than the improvement of cathode properties. In the case of amorphous solid electrolyte, Li conductivity is improved by high Li concentration in unit volume. On the other hand, Li diffusion can readily occur in Li-excess disordered materials, since the Li diffusion energy in such materials, which have a locally Li-rich environment, is decreased as the result of the much weaker electrostatic repulsion [10]. In our previous report [11], we have also elucidated that this theory can be effective against not only Li-excess disordered materials but also inverse spinel materials. Therefore, we expected that the reversible capacity of LiF-excess amorphous samples was higher than that of xLiF-FeSO4 (x = 1.0), which has the same chemical composition as LiFeSO4F. Fig. 4 (a) shows the initial and second charge-discharge curves for tavorite-type and triplite-type LiFeSO4F. The initial charge-discharge capacities were 122/132 mAh g−1 (tavorite type) and 111/ 116 mAh g−1 (triplite type), and the average voltages of the crystalline LiFeSO4F were 3.6 V (tavorite type) and 3.7 V (triplite type). These results were in good agreement with reported values in literatures [3,4]. High operating voltage of these cathode materials occurred by inductive effect of SO4 anion. In particular, the higher operating voltage of triplite-LiFeSO4 may be caused by Li/Fe disordering at octahedral site and the local structure of iron which are octahedrally coordinated by four oxygen and two fluorine. In addition, the overpotential of both crystalline samples was quite small as shown in Fig. 4 (b). On the other hand, the triplite LiFeSO4F prepared by dry ball-milling method has overpotential of about 0.5 V [8], but in our case, it is only 0.2 V. Therefore, we believe that the Li diffusion in the obtained amorphous xLiF-FeSO4 was improved by a locally Li-rich environment. Fig. 5 (a) summarizes the cyclability of the obtained amorphous and crystalline samples at a rate of 0.2 mA cm−2 between 2.5 and 4.3 V. The capacity retentions of the xLiF-FeSO4, tavorite-type, and triplite-type cathodes were 101% (x = 1.0), 100% (x = 1.3), 101% (tavorite type), and 100% (triplite type) even after 30 cycles, respectively. All samples showed excellent cyclability. The rate capabilities of the amorphous xLiF-FeSO4 (x = 1.0, 1.3), tavorite type, and triplite type between 2.5 and 4.3 V (Fig. 5(b)) were investigated with various currents ranging from
4. Conclusions Mechanical co-milling of LiF and nonhydrate FeSO4 successfully gave amorphous xLiF-FeSO4 (1 ≤ x ≤ 2) systems. In addition, we elucidated that tavorite and triplite-type LiFeSO4F can be separately obtained by heating amorphous LiF-FeSO4 at 280 °C and 350 °C, respectively. The amorphous xLiF-FeSO4 cathode has a small overpotential and an average voltage of 3.5 V. Moreover, the xLiF-FeSO4 (x = 1.3) cathode had the highest rechargeable capacity among the amorphous xLiF-FeSO4 series (1 < x < 2), as well as excellent cyclability and rate capability. Although the generally amorphous cathode did not have good electrochemical properties, the Li-excess amorphous cathode can overcome this drawback by percolating the Li network in a locally Lirich environment. References [1] A.K. Padhi, K.S. Nanjundaswamy, C. Masquelier, S. Okada, J.B. Goodenough, J. Electrochem. Soc. 144 (1997) 1609. [2] L. Lander, M. Reynaud, G. Rousse, M.T. Sougrati, C.L. Robert, R.J. Messinger, M. Deschamps, J.-M. Tarascon, Chem. Mater. 26 (2014) 4178. [3] N. Recham, J.-N. Chotard, L. Dupont, C. Delacourt, W. Walker, M. Armand, J.M. Tarascon, Nat. Mater. 9 (2010) 68. [4] P. Barpanda, M. Ati, B.C. Melot, G. Rousse, J.-N. Chotard, M.-L. Doublet, M.T. Sougrati, S.A. Corr, J.-C. Jumas, J.-M. Tarascon, Nat. Mater. 10 (2011) 772. [5] L. Liu, B. Shang, X.-J. Huang, Prog. Nat. Sci. 21 (2011) 211. [6] L. Zhang, J.-M. Tarascon, M.T. Sougrati, G. Rousse, G. Chen, J. Mater. Chem. A 3 (2015) 16988. [7] P. Barpanda, N. Recham, J.-N. Chotard, K. Djellab, W. Walker, M. Armand, J.M. Tarascon, J. Mater. Chem. A 20 (2010) 1659. [8] M. Ati, M. Sathiya, S. Boulineau, M. Reynaud, A. Abakumov, G. Rousse, B. Melot, G.V. Tendeloo, J.-M. Tarascon, J. Am. Chem. Soc. 134 (2012) 18380. [9] M. Kim, Y. Jung, B. Kang, J. Mater. Chem. A 3 (2015) 7583. [10] J. Lee, A. Urban, X. Li, D. Su, G. Hautier, G. Ceder, Science 343 (2014) 519. [11] A. Kitajou, J. Yoshida, S. Nakanishi, Y. Matsuda, R. Kanno, T. Okajima, S. Okada, J. Power Sources 302 (2016) 204.
51