A silica-based gel electrolyte system for improving the cycle performance of LiFePO₄ batteries in an aqueous medium

Abstract

LiFePO₄ based aqueous lithium batteries using aqueous electrolytes suffer from poor cycling performance. This is mainly caused by the Fe dissolution and Li loss generated from the effects of water. In this paper, fumed silica based gel electrolytes were prepared and optimized for improving the cycling performance of the LiFePO₄ electrodes owing to superior stability and comparable ionic conductivity. It was manifested that the Zn/LiFePO₄ cells using this homogenous gel electrolyte showed stable charge/discharge voltage profiles and excellent cycling performance at room temperature. The dissolution of Fe and the loss of Li in this electrolyte is significantly suppressed. These superior performances could endow this gel electrolyte as a promising alternative to aqueous electrolyte systems in the LiFePO₄ batteries at room temperatures.

---

1. Introduction

Lithium transition metal phosphates with olivine structures have received a considerable amount of attention as promising alternative cathode materials for rechargeable lithium ion batteries due to their high energy density, low cost, safety, and chemical stability, which has resulted in the wide use of LiFePO₄ as a cathode material.^1–3 But, organic solvents may be toxic and/or flammable, and various attempts have therefore been made to switch from nonaqueous to aqueous-based systems.^4–9 Manickam et al.¹⁰ carried out a study of the electrochemistry and its surface characterization of LiFePO₄ in aqueous lithium hydroxide electrolyte with metallic zinc as the counter electrode. The Zn|LiOH|FePO₄ battery provided only a discharge capacity of 65 mA h g⁻¹. No cycling performance was reported. Huang et al.¹¹ investigated the kinetics of electrode processes of LiFePO₄ in saturated solution. Milica et al.¹² synthesized the LiFePO₄/C composite as a Li-ion intercalation material in aqueous solutions. In saturated aqueous solution, the initial discharge capacity reached 106 mA h g⁻¹ and the capacity remained 80% of initial value upon 120 charging/discharging cycles. Luo et al.¹³ prepared LiFePO₄ by a sol–gel process. By eliminating oxygen and adjusting the pH values of the electrolyte, the LiTi₂(PO₄)₃/Li₂SO₄/LiFePO₄ aqueous lithium-ion batteries exhibited good stability with capacity retention over 90% after 1000 cycles at a current rate of 6C in pH 13 aqueous electrolyte. But, at a low current rate of C/8, the capacity remained 85% of initial value upon only 50 cycles. Recently, a Zn–LiFePO₄ hybrid secondary battery was proposed by Zhang et al. This battery combined the strong points of Zn–air battery, lithium-ion battery and redox-flow cell. This system can offer an output voltage about 1.2 V and satisfying rate performance above 20C. Also, no prolonged cycling performance was presented.

It has been reported that prolonged exposure of LiFePO₄ to oxygen or water results in Li loss and in increase in Fe(III) content.^5,6,9 Porcher et al.⁵ investigated the stability of LiFePO₄ in aqueous-based systems from an electrochemical perspective, and reported that the surface of LiFePO₄ was changed upon contact with water and a Li₃PO₄ passive layer might be formed. Lee et al.¹⁴ found that the dissolution of Li and Fe ions in aqueous processing yields a depleted powder that has a lower rate capability for LiFePO₄ cathodes.

It is well known that the valve regulated lead acid (VRLA) batteries with a gel electrolyte have excellent performance in several fields.^10,15–18 Gel electrolyte is prepared by mixing a gelling agent with sulphuric acid solution. To improve cycling capability of LiFePO₄ based electrodes in aqueous medium, recently we prepared a new gel electrolyte containing lithium and zinc salts to prevent the dissolution of Li and Fe ions by using fumed silica as a gelling agent. In addition, the gel electrolyte could play a positive role in preventing the generation of Zn dendrite.¹⁹ So, it was anticipated to improve the cycle performance of LiFePO₄ cathodes. In order to explore the feasibility of using fumed silica based gel electrolyte in LiFePO₄ batteries, in this article, the key factors affecting the performance of gel electrolytes were discussed and the composition was optimized. Further, the cyclic capability and the rate discharge performance of LiFePO₄/Zn cells using fumed silica based gel electrolyte were investigated in comparison with those of aqueous electrolyte at room temperatures.

2. Experimental

Commercial LiFePO₄ powder was obtained from Tianjin (Tianjin Stl Energy Technology Co., Ltd) and used without further purification. The working electrode was fabricated by compressing a mixture of the active materials (LiFePO₄), the conductive material (acetylene black), and the binder (polytetrafluoroethylene, PTFE) in a weight ratio of LiFePO₄ : CB : PTFE = 85 : 10 : 5 onto a stainless steel grid by hydraulic machine at 20 MPa. The electrodes were made into the form of square typically (1 cm × 1 cm), and then dried at 120 °C for 1 h. The weight of active material is typically in the range 3–5 mg for each electrode sample. Li₂SO₄ + ZnSO₄ aqueous electrolyte solution was prepared by dissolving Li₂SO₄ and ZnSO₄ in distilled water. And then the pH was adjusted by H₂SO₄, HNO₃ or LiOH. The fumed silica involved in this work were Aerosil 200 (produced by Degussa Co., Germany denoted as A200). The gel electrolyte was prepared by making water-based electrolyte (Li₂SO₄ and ZnSO₄) firstly and then mixing fumed silica and Li₂SO₄ + ZnSO₄ aqueous solutions. The mixture was dispersed in a homogenizer at a high stirring rate to form a colloidal solution. Electrochemical testing was performed after gelation of the colloidal solution.

Cyclic voltammograms, electrochemical impedance spectroscopy (EIS) and charge–discharge experiments were obtained by using a two-electrode beaker cell, which was assembled with a LiFePO₄ cathode and a Zn anode. The active mass of the LiFePO₄ in a piece of cathode is about 5 mg cm⁻² while that of the zinc anode is excessive. CV curves were conducted using an electrochemical working station of CHI608D (Chenhua, Shanghai). The charge/discharge profiles, C-rate capability and cycling ability of cells were recorded on a LAND battery test system. The total ionic conductivity was tested by impedance with graphite blocking electrodes and the electronic conductivity of the electrode was tested by RTS-8 four probe tester (Guangzhou 4-probe technology Co., Ltd. China) the galvanostatic charge/discharge behaviour of Zn–LiFePO₄ cells in aqueous and gel electrolytes were conducted over the range of 0.9–1.6 V. Electrochemical impedance measurements were taken at the open-circuit potential (OCP) of 1.2 V in the frequency range 20 KHz to 0.1 Hz with a Solarton instrument Model 1287 electrochemical interface, and the amplitude of the AC perturbation was 5 mV. All electrochemical measurements were performed at ambient temperature.

3. Results and discussion

3.1 Optimization of gel electrolyte

3.1.1 Effects of fumed silica concentration. In gel batteries, fumed silica thickens and absorbs the electrolyte solution. The concentration of fumed silica is a key factor affecting the gelling properties. Fig. 1 shows the effects of fumed silica content on the dispersion stability of gel electrolyte. It is observed that after 20 days, when the content of fumed silica added is more than 4.5%, no lay separation occurred and a uniform gel aqueous electrolyte was formed. But, the more the amount of fumed silica added, the lower the ionic conductivity of gel electrolyte is. It has been confirmed that with increasing silica content, the gelling time shortens and the structure of the gel becomes more compact. Consequently, the diffusion of reactants is inhibited, and the electrode capacity tends to diminish. In addition, the ohmic resistance and the charge-transfer resistance increase with higher silica content. Thus, according to the electrochemical properties of LiFePO₄ in gel electrolyte, the concentration of fumed silica is further optimized.

Fig. 1 Effects of the fumed silica content on the stability of gel electrolyte. Silica content: (1) 1%; (2) 2.8%; (3) 3.2%; (4) 4.2%; (5) 4.5%; (6) 5.3%.

Fig. 2 shows the CV curves and rate capability of the LiFePO₄ electrode in 0.5 M Li₂SO₄ + 2 M ZnSO₄ gel electrolytes prepared with different silica contents. With optimal dispersion, the response current of gel electrolytes prepared with 4.5 and 5.3 wt% silica are almost equal. However, when the silica content is increased to 6 wt%, a significant decrease in the oxidation and reduction peak current is presented. It can be more clearly seen that the LiFePO₄ electrode in gel electrolytes with 4.5 and 5.3 wt% silica exhibits much better rate dischargeability than that in gel electrolytes with 6 wt% silica, which is likely related to the ionic conductivities of the different gel electrolytes.

Fig. 2 CV curves (A) and rate capability (B) of the LiFePO₄ electrode in 0.5 M Li₂SO₄ + 2 M ZnSO₄ gel electrolytes prepared with different silica contents: (a) 4.5 wt%; (b) 5.3 wt%; (c) 6.0 wt%.

Since the dispersion of the fumed silica plays an important role in the electrochemical properties of LiFePO₄ electrodes in gel electrolytes, CV and EIS were employed to evaluate the effect of agitation time on the electrochemical performance of the gels. CV results for gel electrolytes prepared under different agitation time are shown in Fig. 3A and B. Initially, the redox peak currents increase with increasing agitation time, reaching a maximum at about 120 min before decreasing. It can be concluded that a lower or higher agitation time will lead to a sharp decrease in the peak current of the LiFePO₄ electrode.

Fig. 3 CV results of the LiFePO₄ electrode in gel electrolytes (4.5 wt% fumed silica, room temperature of 25 °C) prepared with different agitation times: (A) CV curves and (B) relationships between agitation time and redox peak currents of the LiFePO₄ electrode derived from (A).

The impedance spectra (performed at 1.2 V) for gel electrolyte prepared with different agitation times is shown in Fig. 4A. The results can be fitted well by the equivalent circuit included (see Fig. 4B). R₁ represents the ohmic resistance, made up of the resistance of the electrolyte, and the resistance of the electrical connections to the electrode. R₂ is the charge-transfer resistance of the rate-controlling electrochemical reaction of the lithium-ion intercalation and deintercalation process. CPE is the double-layer capacitance. It can be seen that the solution resistance (R₁) and charge-transfer resistance (R₂) of the gel reach a minimum with the optimal agitation time (120 min) (Fig. 4B). Therefore, the EIS results are consistent with the results obtained from the CV study.

Fig. 4 Impedance spectra results of the LiFePO₄ electrode in gel electrolytes (4.5 wt% silica) prepared with different agitation times and (B) relationships between solution resistance (R₁) or charge-transfer resistance (R₂) and agitation time derived from (A).

3.1.2 Effects of electrolytic salts. Besides the gelling agent, electrolytic salts such as Li₂SO₄ and ZnSO₄ also exert great effects on the performance of gel electrolytes. Table 1 summarizes the composition of the gel electrolytes. The net content of silica in each gel is 4.5 percent of total weight, and the concentration of Li₂SO₄ and ZnSO₄ was as follows: 1 M/0.5 M (Sample 1#), 1 M/1 M (Sample 2#), 1 M/2 M (Sample 3#), 0.5 M/0.5 M (Sample 4#), 0.5 M/1 M (Sample 5#), 0.5 M/2 M (Sample 6#). The mixture was stirred at a high speed in a homogenizer to form a colloidal solution.

Table 1 Gel electrolyte composition

Sample	1#	2#	3#	4#	5#	6#
Li2SO4/M	1	1	1	0.5	0.5	0.5
ZnSO4/M	0.5	1	2	0.5	1	2

---

Fig. 5 presents the CV curves of the LiFePO₄ electrode in gel electrolytes with varying concentration of Li₂SO₄ and ZnSO₄. As can be seen, compared with Li₂SO₄, a high concentration of ZnSO₄ is more in favor of improving the kinetics of the LiFePO₄ electrode. As a result, the corresponding peak currents of the gel electrolyte (3# and 6#) are much higher than the other gel electrolytes. And compared with 3# gel, 6# gel exhibits a lower oxidation peak potential, implying that the polarization of the LiFePO₄ electrode is reduced further. This suggests that the 6# gel electrolyte system will provide the excellent kinetics and rate performance for the LiFePO₄ electrode.

Fig. 5 Cyclic voltammograms of the LiFePO₄ electrode in gel electrolytes with varying concentration of Li₂SO₄ and ZnSO₄ at the 5th cycle (v = 10 mV s⁻¹).

Fig. 6 shows the electrochemical impedance spectra of the gel electrolyte with varying concentration of Li₂SO₄ and ZnSO₄ obtained at open circuit potential. The results can be fitted well by the equivalent circuit of Fig. 4B. The plots for the six gel compositions are similar and all of them exhibit a semicircle at high frequency and a long slant in the middle and low-frequency regions. This indicates that solid-state diffusion of lithium ion is a rate determining step. The slant for the 6# gel electrolyte is much shorter than the other gel electrolytes. It also can be seen from the fitting results shown in Table 2 that unequal concentration of Li₂SO₄ and ZnSO₄ (0.5 : 2) generate the lowest solution resistance and charge-transfer resistance. This is generally consistent with rate capability testing below.

Fig. 6 EIS for gel electrolyte prepared with varying concentration of Li₂SO₄ and ZnSO₄ at open circuit potential.

Table 2 Summaries of the fitting results of ohmic resistances (R₁) and charge-transfer resistances for Li⁺ intercalation (R₂)

Gel electrolytes	1#	2#	3#
R1	2.49	2.54	2.85
R2	17.08	10.81	5.03
Gel electrolytes	4#	5#	6#
R1	2.85	2.76	2.30
R2	13.99	11.09	3.68

---

Fig. 7 shows the discharge specific capacities of the LiFePO₄ electrode in different gel electrolytes at different charge/discharge current rates. In the current rate of 0.2–4C, the specific capacities of the electrodes in different gel electrolytes decrease with current rate increasing, but show different fading magnitudes. When the current rate is increased from 0.2C to 4C, the specific capacities of the electrodes in 1#, 2#, 3#, 4#, 5# and 6# gel electrolytes are decreased from 135.2, 136.6, 134.5, 136.1, 134.2 and 136.2 mA h g⁻¹ to 60.2, 59.2, 63.1, 57.4, 59.9 and 64.8 mA h g⁻¹, i.e., decreased by 55.5%, 56.7, 53.1%, 57.8%, 55.4% and 52.4%, respectively. The 6# gel electrolyte has the highest rate capabilities and then the 3# gel electrolyte, which is in agreement with the ac impedance results in Fig. 6 and Table 2. That is to say, the rate discharge ability decreases with the electrode impedance increasing. The above results indicate that the LiFePO₄ electrode in 3# or in 6# gel electrolyte has higher specific capacity and discharge potential as well as better rate dischargeability, which is likely related to the Li⁺ ions activities and Zn²⁺ ions conductivities of the different gel electrolytes.

Fig. 7 Discharge specific capacities of LiFePO₄ electrodes in different gel electrolytes at different current rates.

3.2 Gel batteries performance

The 6# gel electrolyte is employed to investigate the LiFePO₄ batteries performance with comparison to aqueous electrolytes. The variation in electrode stability during cycling was first measured using cyclic voltammetry (CV). Herein, the cells were cycled at a scan rate of 1 mV s⁻¹ under a voltage range of 0.9–1.6 V. Fig. 8(A and B) showed the cyclic voltammograms of the LiFePO₄ cells using aqueous solution and gel electrolytes after the 1st and 110th cycle at room temperature. Similar to the previous results, one pair of major redox peaks were observed. In cyclic voltammograms, the peak currents and potential separations of the redox were known to indicate the kinetics, wherein a lower peak current and larger potential separation represented stronger electrode polarization. In the case of aqueous electrolytes, the peak currents decreased from 9.34 mA (the 1st cycle) to 7.47 mA (after the 110th cycle). In comparison, the peak currents for the gel electrolyte were reduced from 8.5 mA to 8.37 mA. A quantitative comparison of the above results confirmed that it was effective in preventing the Fe dissolution and Li loss from LiFePO₄ by employing gel electrolytes, indicative of improving the cycle stability of LiFePO₄ electrodes.

Fig. 8 Cyclic voltammograms of the LiFePO₄ cells using aqueous solution (A) and gel electrolytes (B) after the 1st and 110th cycle at room temperature (at a scan rate of 1 mV s⁻¹ under a voltage range of 0.9–1.6 V).

The cycling performance of the Zn/LiFePO₄ cells using aqueous and gel electrolytes at room temperature with various current rates were depicted in Fig. 9. As we can see from Fig. 9A, though both cells deliver similar initial discharge capacity and exhibit regular graded capacity change in the first 100 cycles, in the subsequent cycles the capacity in aqueous electrolyte drops more quickly than that in gel electrolyte, especially in higher current rates. The cycling data in Fig. 9A is also plotted as capacity retention versus time shown in Fig. 9B. The cell shows the same capacity fading tendency at both high and low current rates.

Fig. 9 Cycling performance of LiFePO₄ electrodes in aqueous and gel electrolytes at room temperature with various current rates. (A) The discharge capacity versus cycle number, (B) capacity retention versus cycle number. Solid symbol is in the gel electrolyte and hollow symbol is in the aqueous electrolyte, (C) charge and discharge curves of the Zn/LiFePO₄ cells using aqueous and gel electrolytes at the current rate of 0.5C.

The capacity retention in the gel electrolyte is more than 75% after 180 cycles with various current rates, while in the aqueous electrolyte, it is only 60%. Since the battery performance is mainly dependent on the LiFePO₄, it implies that the capacity fading of LiFePO₄ in gel electrolytes is effectively prevented. The corresponding charge/discharge curves of the Zn/LiFePO₄ cells using aqueous and gel electrolytes are presented in Fig. 9C. Both cells exhibited almost overlapped charge/discharge curves and the charge/discharge plateaus at about 1.24 V vs. Zn/Zn²⁺ and 1.13 V vs. Zn/Zn²⁺ respectively in the first cycle. But, in the 180th cycle, the capacity of LiFePO₄ in the aqueous electrolyte is rather lower than that in the gel electrolyte.

In an effort to gain deep insight into the positive influence of the gel electrolyte on the cycling performance of LiFePO₄ electrodes, the AC impedance spectra of the cells after the 1st and 180th cycle at room temperature were depicted in Fig. 10. It was shown that after 180 cycles, the cell impedance in the aqueous electrolyte increased more significantly than that in the gel electrolyte. This indicated that the capacity fading of LiFePO₄ in aqueous electrolyte during cycling was closely associated with the continuous increase of cell resistance, which might result from the gradual dissolution of Fe and the loss of Li. This is demonstrated from Table 3, the content of dissolved Fe and Li loss in the aqueous electrolyte is nearly twice more than that in the gel electrolyte. Comparatively, although the initial cell impedance for the gel electrolyte was slightly higher than that of the aqueous electrolyte, the growth of cell impedance for the gel electrolyte was significantly retarded. This demonstrated that the use of fumed silica-based gel electrolyte may effectively alleviate the Fe dissolution and suppress the formation of a passive layer on the LiFePO₄ surface during charge/discharge cycling.

Fig. 10 AC impedance spectra of the cells using gel electrolyte (A) and aqueous solution based electrolyte (B) after the 1st and 180th charge/discharge cycle at room temperature.

Table 3 The content of dissolved Fe(III) and Li loss in the aqueous and gel electrolytes after 180 cycles^a

	Fe	Li
a The volume of aqueous and gel electrolytes was 10 ml.
Aqueous electrolytes/ppm	48.4	6.1
Gel electrolytes/ppm	21	2.8

---

3.2.1 Comparison of rate capability. The cyclability of the LiFePO₄ samples with various applying current at room temperature in aqueous and gel electrolytes was shown in Fig. 11. As can be seen, the initial capacity (125 mA h g⁻¹) was attained when the cell discharge at 0.2C rate in aqueous and gel electrolytes. At 1C, the discharge capacity in the gel electrolyte is still slightly higher than that in aqueous electrolyte. But, after that, the current rate the larger, the higher the discharge capacity of LiFePO₄ is in the aqueous electrolyte. This may be related to the ionic conductivity of the electrolyte. We measured the total ionic conductivity of the liquid and gelled electrolytes by impedance with blocking electrodes, which are 0.43 and 0.37 S cm⁻², respectively. As can be seen, it caused LiFePO₄ with certain sacrifice on the rate capability. When the current rate is higher than 2C, a significant decrease in the discharge capacity of LiFePO₄ just can be observed in the gel electrolyte.

Fig. 11 Cycle stability of the LiFePO₄ samples with various applying current in aqueous and gel electrolytes.

4. Conclusion

Fumed silica based gel electrolytes were prepared and optimized in aqueous medium of pH 5. The Zn/LiFePO₄ cells using homogenous gel electrolytes demonstrated stable charge/discharge voltage profiles and excellent cycling performance, though the rate capability is somewhat lowered. Moreover, fumed silica based gel electrolyte alleviated the growth of cell impedance and the Fe dissolution, lithium loss problems. The superior cycling performances could endow this class of fumed silica based gel electrolyte a very promising alternative to state of the art liquid electrolyte system in the LiFePO₄ lithium batteries.

Acknowledgements

The acknowledgements come at the end of an article after the conclusions and before the notes and references.

Notes and references

1. J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
2. P. G. Bruce, B. Scrosati and J. M. Tarascon, Angew. Chem., 2008, 47, 2930–2946.
3. Y. Wang, J. Wang, J. Yang and Y. Nuli, Adv. Funct. Mater., 2006, 16, 2135–2140.
4. J. H. Lee, U. Paik, V. A. Hackley and Y. M. Choi, J. Electrochem. Soc., 2005, 152, A1763.
5. W. Porcher, P. Moreau, B. Lestriez, S. Jouanneau and D. Guyomard, Electrochem. Solid-State Lett., 2008, 11, A4.
6. J. F. d. r. Martin, A. Yamada, G. Kobayashi, S.-i. Nishimura, R. Kanno, D. Guyomard and N. Dupré, Electrochem. Solid-State Lett., 2008, 11, A12.
7. J. H. Lee, J. S. Kim, Y. C. Kim, D. S. Zang and U. Paik, Ultramicroscopy, 2008, 108, 1256–1259.
8. W. Porcher, B. Lestriez, S. Jouanneau and D. Guyomard, J. Electrochem. Soc., 2009, 156, A133.
9. Y. W. Y. Denis, K. Donoue, T. Kadohata, T. Murata, S. Matsuta and S. Fujitani, J. Electrochem. Soc., 2008, 155, A526–A530.
10. M. Manickam, P. Singh, S. Thurgate and K. Prince, J. Power Sources, 2006, 158, 646–649.
11. K. L. Huang, S. Yang, S. Q. Liu and H. B. Wang, Acta Phys.-Chim. Sin., 2007, 23, 129–133.
12. M. Vujković, I. Stojković, N. Cvjetićanin and S. Mentus, Electrochim. Acta, 2013, 92, 248–256.
13. J. Y. Luo, W. J. Cui, P. He and Y. Y. Xia, Nat. Chem., 2010, 2, 760–765.
14. J. H. Lee, H. H. Kim, G. S. Kim, D. S. Zang, Y. M. Choi, H. Kim, D. K. Yi, W. M. Sigmund and U. Paik, J. Phys. Chem. C, 2010, 114, 4466–4472.
15. S. Hyub Oh, M. Kim, J. Bok Lee and H. L. Lee, Bull. Korean Chem. Soc., 2002, 23, 75–80.
16. S. K. Martha, B. Hariprakash, S. A. Gaffoor and A. K. Shukla, Bull. Mater. Sci., 2003, 26, 465–469.
17. J. C. Hernández, M. L. Soria, M. González, E. García-Quismondo, A. Muñoz and F. Trinidad, J. Power Sources, 2006, 162, 851–863.
18. H. Zhang, X. Wu, T. Yang, S. Liang and X. Yang, Chem. Commun., 2013, 49, 9977–9979.
19. B. Qin, Z. Liu, G. Ding, Y. Duan, C. Zhang and G. Cui, Electrochim. Acta, 2014, 141, 167–172.

---