The Li_xBp(DME)_9.65 solutions were prepared in a glove box filled with argon. Different amounts of lithium metal (China Energy Lithium Co. Ltd, China), biphenyl (alfa Aesar, China) and DME (alfa Aesar, China) were prepared according to “x” of the Li_xBp(DME)_9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions. Then, biphenyl was dissolved in DME by magnetic stirring for 2 h at room temperature in a glass bottle. In order to increase the dissolution rate, the lithium metal could be cut into small pieces. Finally, the lithium metal pieces were put into the biphenyl DME solution with string for 2 h at room temperature.

The Li_xBp(DME)_9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions were sealed in a glass bottle for the conductivity measurement. The cell had two stainless steel electrodes with an area of 0.5 cm² and a distance of 0.8 cm. An electrochemical workstation (Zahner IM6e) was used for impedance spectroscopy measurements. The data was recorded in the frequency range from 100 Hz to 8.0 MHz by applying an alternating voltage of 5 mV at different temperature (0–40 °C). The conductivity cell was calibrated by a standard 0.1 M potassium chloride solution at 25 °C. For the ITIC measurements, the voltage was 0.1 V between two electrodes and the current was recorded for 30 minutes at 20 °C.

The LFP (BTR New Energy Material Co. Ltd, China) electrode was prepared by a slurry coating procedure. The slurry was composed of LFP as the active material, carbon black (alfa) as the conductivity additive and polyvinylidene fluoride (alfa) as the binder in a weight ratio of 80:10:10 dissolved in N-methyl pyrrolidone (alfa). Then the slurry was spread uniformly onto an aluminum foil and dried in a vacuum oven at 120 °C for 6 h. For the anodes, carboxymethylcellulose (CMC)/styrene-butadiene rubber (SBR) was chosen as the binder. The current collector was copper foil. The weight ratio of CMC and SBR was 2:3. The active materials of the anode were hard carbon (BTR New Energy Material Co. Ltd, China), soft carbon (BTR New Energy Material Co. Ltd, China), graphite (BTR New Energy Material Co. Ltd, China), and nano-silicon (alfa), respectively.

For the prelithiation process, the anode electrodes were immersed in the Li_1.00Bp(DME)_9.65 solution for 15 s, 40 s, 1 min, and 5 min, respectively. The electrodes were then washed by dimethyl carbonate to remove the excess biphenyl. The electrochemical measurements of the electrodes were carried out with 2032 coin cells using lithium metal as the counter electrode. The solution of 1.0 M LiPF₆ and vinylene carbonate (2% in volume) in ethylene carbonate and dimethyl carbonate (1:1 in volume) was served as the electrolyte. The cells were galvanostatically charged and discharged at a current density of 100 mA/g between 0.005 V and 2.0 V (vs. Li⁺/Li) on the Land BA2100A battery testing system (Wuhan, China) at room temperature.

For the half liquid battery, a new structure of battery is shown in Fig. 1. The Li_1.00Bp(DME)_9.65 solution was used as the anode and the copper foam as the current collector. We selected the LFP electrode to be the counter cathode, and use the LATP solid state electrolyte pellet to separate the cathode and anode.^[28] The solution of 1.0 M LiPF₆ in ethylene carbonate and dimethyl carbonate (1:1 in volume) was served as the electrolyte for the cathode side. The battery was galvanostatically charged and discharged at a current density of 50 mA/g between 2.20 V and 3.75 V (vs. Li⁺/Li) at room temperature by a Land BA2100A battery testing system (Wuhan, China).

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (color online) Structure of the flow battery. The Blue part is stainless steel, Grey parts are Teflon, the Red one is Li1.00Bp(DME)9.65 solution, the White part is LATP, Yellow is the LFP electrode and Green is the electrolyte.

The current versus time curve at 0.1 V (20 °C) between two stainless steel electrodes of five kinds of Li_xBp(DME)_9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions are shown in Fig. 2(a). When , the current decreases quickly and stabilizes around a current value. For others, the current does not change with time obviously. The almost invariant current indicates low ionic conductivity for x below 1.00, which is due to the electronic conductor and ionic insulator property of stainless steel electrodes. The rapid decline of the current indicates the existence of ionic transport. The electronic conductivity always exists from the final value of the current of these curves.

Fig. 2.

Figure Option ViewDownloadNew Window

Fig. 2. (color online) (a) The current versus time of LixBp(DME)9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions at 20 °C during the 0.1 V step. (b) The current versus time of Li1.00Bp(DME)9.65 solution at 20 °C during the 0.1 V step, experimental data (blue) and fitting line (red). (c) The fitting result of electronic conductivities and ionic conductivities of LixBp(DME)9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions at 20 °C.

The ionic conductivities and electronic conductivities of the Li_xBp(DME)_9.65 solutions are achieved by fitting the curves shown in Fig. 2(b)^[29,30] with the following formula:

The fitting result is shown in Fig. 2(c). The electronic conductivity enhances as the lithium concentration increases up to 7.2 mS/cm. However, the rising speed slows down gradually. The ionic conductivity enhances with the increasing of lithium concentration in this work. It is noticed that the ionic conductivity is several orders of magnitude lower than the electronic conductivity for the Li_0.50Bp(DME)_9.65 and Li_0.25Bp(DME)_9.65. It reveals that the electrons migrate more easily in the solution than in ions. This is an amazing result since nearly all reported solutions have negligible electronic conductivity at room temperature. With the lithium concentration increasing, the ionic conductivity is close to the electronic conductivity. Compared with the electronic conductivity, the ionic conductivity always shows a lower value when the lithium concentration is lower than Li_2.00Bp(DME)_9.65.

The impedance spectrum of Li_xBp(DME)_9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions at 20 °C are shown in Fig. 3(a). The conductivities are calculated by the following formula:

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (color online) (a) The impedance spectrum of LixBp(DME)9.65 (x = 0.25, 0.50, 1.00, 1.50, 2.00) solutions at 20 °C. The inset shows the local amplification. (b) The total conductivities of LixBp(DME)9.65 solutions from the impedance spectroscopy. (c) Temperature versus total conductivities of LixBp(DME)9.65 solutions.

The versus lithium concentration is shown in Fig. 3(b). The enhances as the lithium concentration increases up to 10.7 mS/cm. The rising speed reduces obviously as “x” is higher than 1.00. With the increasing of lithium content increasing, the conductions of ions and electrons are influenced by the interaction between them. The relation between total conductivity and temperature is shown in Fig. 3(c). The of Li_xBp(DME)_9.65 solutions enhance in higher temperature. A perfect linear relationship is obtained. This means that the system obeys the Arrhenius law, which allows us to estimate the activation energy , ). The higher ionic conductivity implies more correlation between activation energy and temperature. Compared with LiPF₆-PC (0.12 eV), the lower is due to the low viscosity of DME.^[31,32] The slope of fitting lines is much lower for the Li_0.50Bp(DME)_9.65 and Li_0.25Bp(DME)_9.65 compared with the higher lithium concentration ones. This is reasonable that the ionic conductivity in the high concentration system is more significant than that in the dilute solution system.

The charge–discharge curve of the half liquid battery at the first cycle is shown in Fig. 4(a). The specific capacity of LFP is 130 mAh/g, which is lower than 160 mAh/g in normal Li-ion batteries, due to the large cell polarization (0.3 V) and low charge cut-off voltage 3.75 V. The first coulombic efficiency is 95%, which is obviously higher than that of the graphite-based anode battery due to the presence of the SEI formation. The charge and discharge potential platforms of the full battery are 3.3 V and 3.0 V respectively, and the LFP versus Li⁺/Li is 3.45 V.^[33] Therefore, the potential of Li_1.00Bp(DME)_9.65 solution versus Li⁺/Li is about 0.3 V. This illustrates that the full cell has high voltage and high energy density. The polarization between the charge curve and discharge curve is as large as 0.3 V. One reason for the large polarization is the low ionic conductivity of the LATP solid electrolyte pellet.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. The performance of the LFP-Li1.00Bp(DME)9.65 solution battery: (a) Charge–discharge curve of the first cycle; (b) cycling performance.

The cycling performance of the half liquid battery is shown in Fig. 4(b) which does not satisfy the specific capacity decay of 10 mAh/g after 20 cycles. This is due to the reactions between the LATP and the Li_1.00Bp(DME)_9.65 solution. The low redox potential of Li_1.00Bp(DME)_9.65 solution (0.3 V versus Li⁺/Li) indicates strong reducibility. The LATP is not stable below 1.5 V versus Li⁺/Li as is well known.^[28]

As a result of the strong reducibility of Li_1.00Bp(DME)_9.65 solution (0.3 V versus Li⁺/Li), it is a convenient solution agent for prelithiation. Figure 5 shows the electrochemical performance of anode materials with and without lithiation of graphite, nano-silicon, soft carbon and hard carbon. For the graphite, the first coulombic efficiencies are 88.6% (without lithiation) and 89.3% (after lithiation for 40 s) as shown in Fig. 5(a). This result indicates that the presence of Li_1.00Bp(DME)_9.65 solution is difficult to lithiate the graphite anode due to the low lithiation voltage of graphite.^[34] The potential is lower than the Li_1.00Bp(DME)_9.65 solution voltage (0.3 V). For the soft carbon, the first coulombic efficiencies are 86.1% (without lithiation), 94.2% (after lithiation for 15 s), 107.5% (after lithiation for 40 s), 112.7% (after lithiation for 1 min) and 124.9% (after lithiation for 5 min), which is shown in Fig. 5(b). For the hard carbon, the values of the first coulombic efficiency are 74.8% (without lithiation), 91.2% (after lithiation for 15 s), 95.2% (after lithiation for 40 s), 95.8% (after lithiation for 1 min), and 161.7% (after lithiation for 5 min) as shown in Fig. 5(c). For the hard carbon and soft carbon, these results show that their first coulombic efficiencies can be improved obviously. Figure 5(d) tells us that the first coulombic efficiency of nano-silicon has improved less (3.9%) than soft carbon (21.4%) and hard carbon (20.4%) after a 40 s lithiation process. This difference indicates that the kinetic of lithiation into nano-silicon is not as good as hard carbon and soft carbon. Figures 5(e)–5(g) demonstrate that the cycling performance of the nano-silicon, hard carbon and soft carbon do not deteriorate after the lithiation process.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. (color online) Half battery performance of lithiated graphite (a), soft carbon ((b), (e)), hard carbon ((c), (f)), nano-silicon ((d), (g)) electrodes. (a), (b), (c), (d) are the first charge-discharge curves. (e), (f), (g) are the cycling performance.