Large-scale battery systems are essential for efficiently utilizing renewable energy power sources from solar and wind, which can generate electricity only intermittently. The use of lithium-ion batteries to store the generated energy is one solution. A long cycle life is critical for lithium-ion battery when used in these applications; this is different from portable devices which require 1,000 cycles at most. Here we demonstrate a novel co-substituted lithium iron phosphate cathode with estimated 70%-capacity retention of 25,000 cycles. This is found by exploring a wide chemical compositional space using density functional theory calculations. Relative volume change of a compound between fully lithiated and delithiated conditions is used as the descriptor for the cycle life. On the basis of the results of the screening, synthesis of selected materials is targeted. Single-phase samples with the required chemical composition are successfully made by an epoxide-mediated sol-gel method. The optimized materials show excellent cycle-life performance as lithium-ion battery cathodes.

INTRODUCTION LiFePO4 is a commercially successful cathode material due to low costs and high-power energy density. The electrode reaction proceeds divided into two phases, Li-rich phase and Li-poor phase, with a large volume change. We reported the intrinsic high-rate nature that the intermediate phase, which is metastable at room temperature, emerges to moderate the lattice mismatch and gets closer to the single-phase reaction1. Here, we show how the stabilization of the intermediate phase can improve the rate capability using time-resolved in-situ XRD measurements. We also derive general non-equilibrium criteria for two-phase reaction in Li-intercalation compounds using the LiFePO4 system as a specific example. EXPERIMENTAL Undoped LiFePO4 and Li(Fe0.95Zr0.05)(P0.9Si0.1)O4, denoted UL and Z2S respectively, were synthesized in the same manner as reported2. The partially delithiated materials which is Li concentration of x = 0.66 in LixFePO4 were prepared by the chemical reduction. The ex situ X-ray diffraction (XRD) patterns were obtained at various Li concentration and elevated temperature. The electrochemical tests were performed using a liquid electrolyte at room temperature and a molten salt electrolyte at elevated temperature. The time-resolved in situ XRD measurements were performed at BL28XU, SPring-8 with a wavelength of 0.619862(2) Å using a 1D detector, MYTHEN 1K. The data were collected in the 2θ range of 10° to 13° with an exposure time of 1 s. RESULTS AND DISCCUSION The ex situ XRD patterns of Li0.66FePO4 show that the intermediate phase explicitly emerge at 150°C in Z2S whereas 200°C in UL, which indicate the stabilization of the intermediate phase by the reduced lattice volume change. This result is also supported the electrochemical tests at elevated temperature whether or not the two-step voltage plateau could be observed. The ex situ XRD patterns of various Li concentration and the open-circuit voltage measurements show a typical two-phase characteristics in both cathodes. There is no significant difference in phase transition behavior based on an equilibrium system because the intermediate phase is metastable at room temperature. We performed time-resolved in-situ XRD measurements to access the non-equilibrium phase transition under a large overpotential. At a high rate of 10C, the peak shift mainly occurs in Z2S compared with UL. This suggests the quasi-single phase reaction proceeds via the intermediate phase. The reaction kinetics was then evaluated. At higher rates, the differences of both capacities are lager, which indicates the enhancement of the rate capability in Z2S. The stabilization of the intermediate phase by the reduction of the lattice volume change leads to the ability of high-rate cycling. This insight involves new concepts in non-equilibrium thermodynamics, which may apply other two-phase electrode materials. REFERENCES 1. Y. Orikasa, T. Maeda, Y. Koyama, H. Murayama, K. Fukuda, H. Tanida, H. Arai, E. Matsubara, Y. Uchimoto and Z. Ogumi, J. Am. Chem. Soc., 135, 5497 (2013). 2. M. Nishijima, T. Ootani, Y. Kamimura, T. Sueki, S. Esaki, S. Murai, K. Fujita, K. Tanaka, K. Ohira, Y. Koyama and I. Tanaka, Nat. Commun., 5 (2014). 3. A. Van der Ven, K. Garikipati, S. Kim and M. Wagemaker, J. Electrochem. Soc., 156, A949 (2009).

INTRODUCTION Lithium iron phosphate, LiFePO4, becomes conspicuous as a commercially important cathode material due to low cost, high safety, and non-toxic nature. Lithium-ion intercalation proceeds through a two-phase reaction between two compositions close to the endmembers LiFePO4 (LFP-phase) and FePO4 (FP-phase). The first-order phase transformation accompanied with a large volume change of 6.8% hinders moving phase boundaries faster than ever. Co-substituted LiFePO4, represented in Li(Fe1-xZrx)(P1-2xSi2x)O4 or Z2S, which decreases the lattice volume change between two phases, shows six times longer cycle life than undoped LiFePO4 (ref. 1). Here we focus on the Z2S, consisting of moderate two-phase interfaces, and investigate kinetics and mechanisms for the two-phase reaction with reduced lattice mismatch using time-resolved X-ray diffraction. EXPERIMENTAL Undoped LiFePO4 and Li(Fe0.95Zr0.05)(P0.9Si0.1)O4, just called hereafter Z2S, were synthesized in the same manner as reported1. The ex situ X-ray diffraction (XRD) measurements were performed at BL02B2, SPring-8 with a wavelength of 0.699292(4) Å using a Debye-Scherrer camera and an imaging plate detector. The cathode materials for the electrochemical tests were prepared by mixing 80% active material, 10% carbon black, and 10% polyvinylidene fluoride (PVDF) with 1-methyl-2-pyrrolidinone solvent. The composite electrodes were placed in original laminate-type cells in an Ar-filled glovebox with lithium metal as the counter and reference electrodes. LiPF6 (1 M) in a 3:7 volume ratio of ethylene carbonate (EC) and ethyl methyl carbonate (EMC) was used as the electrolyte. The in situ XRD measurements were performed at BL28XU, SPring-8 with a wavelength of 0.619862(2) Å using a 1D detector, MYTHEN 1K. The data were collected in the 2θ range of 10° to 13° with an exposure time of 1 s every 60 s(1C) and 15 s(10C). RESULTS AND DISCCUSION Both Undoped LiFePO4 and Z2S are the same particle size as about 150 nm, indicating independent of nano-sized effects2 and allowing the comparison of the difference of the lattice volume change. Galvanostatic charge/discharge tests were performed. At a low rate of 1C, both of them show the same capacity as 120 mAh/g and the similar shape of charge/discharge curves. At a high rate of 10C, the capacity of Z2S keeps still 90 mAh/g although that of undoped LiFePO4 falls to 75 mAh/g. The difference of both capacities are lager with increasing rates, which indicates the enhancement of the rate capability in Z2S. In order to get information on the static state, the ex-situ XRD and the open-circuit voltage (OCV) measurements were performed. The ex situ XRD patterns show a typical two-phase characteristics in both cathodes as emerged FP-phase peaks from 0.1Li during a charge process. The OCV curves show the flat voltage plateau in a wide range, which also indicates the coexistence of two phases in the reaction. From these results based on an equilibrium system, there is no significant difference in phase transition behavior between undoped LiFePO4 and Z2S. To try to understand what is going on in lithium intercalation reaction indeed and to achieve high power lithium-ion batteries, time-resolved in-situ XRD measurements were performed. At a low rate of 1C, undoped LiFePO4 proceeds via two phase reaction similarly in the results of the ex situ experiments. In Z2S, the peaks belonging to each LFP-phase and FP-phase simply increase or decrease although accompanying with peak broadening. At a high rate of 10C, on the other hand, the peak shift mainly occurs in Z2S compared with undoped LiFePO4. This suggests the quasi-single phase reaction proceeds in Z2S and explains the enhancement of rate capability in Z2S, which is the small lattice volume change. REFERENCES 1. M. Nishijima, T. Ootani, Y. Kamimura, T. Sueki, S. Esaki, S. Murai, K. Fujita, K. Tanaka, K. Ohira, Y. Koyama and I. Tanaka, Nat. Commun., 5 (2014). 2. N. Meethong, H. Y. S. Huang, W. C. Carter and Y. M. Chiang, Electrochemical and Solid State Letters, 10, A134 (2007).