Performance and potential problems of high power density zinc–nickel single flow batteries

Abstract

High power density with high efficiency can facilitate rapid charge–discharge and reduce the cost of zinc–nickel single flow batteries, and therefore it is of significant technological importance. In this paper, the battery performance and potential problems have been investigated at high current density up to 300 mA cm⁻², which is the highest current density that has ever been obtained. The results show that coulombic efficiency first increases and then decreases with the current density increasing due to the non-uniform distribution of electrode potential and side reactions. The positive electrode deeply discharges and zinc accumulates on the negative electrode at the end of the discharging process at a high current density. The morphologies of the deposited zinc vary from smooth, spongy to dendrite with the increasing current density. Moreover, the positive polarization is a critical obstacle to improve the performance of zinc–nickel single flow batteries at a high current density. Based on these findings, we point out the remaining issues and struggling directions enabling high power density ZNBs without the substantial loss of cycle life.

---

1. Introduction

Zinc–nickel single flow battery (ZNB) is a promising energy storage device for improving the reliability and overall use of renewable energy due to its advantages of low cost, high efficiency, long cycle life, high open circle potential (1.705 V) and high energy density.^1–3 The anode is a zinc electrode with high electrode potential (Zn(OH)₄²⁻ + 2e = Zn + 4OH⁻, −1.215 V vs. SHE, standard hydrogen electrode) and high specific capacity (820 mA h g⁻¹), which is widely used in zinc-based batteries.^4,5 The cathode is a nickel hydroxide/nickel oxyhydroxide electrode (Ni(OH)₂ − e + OH⁻ = NiOOH + H₂O, 0.49 V vs. SHE) that is diffusively employed in Ni–Cd and Ni-MH batteries.^6–8 Moreover, no membrane is needed between positive and negative electrodes, which simplifies the structure and reduces the cost compared to the traditional two flowing electrolyte batteries. Different from the traditional zinc–nickel batteries, the flowing electrolyte enhances proton transport at the cathode and inhibits zinc dendrite at the anode in ZNBs, which remarkably improves the energy efficiency and cycle life.

The energy may be generated or consumed in a short time, which requires a battery with high power density. In addition, the stack with high power density can reduce the cost of direct materials, and thus reduce the cost of the flow batteries.⁹ However, since a long time, ZNB has suffered from large polarization that resulted in low operating current density (below 20 mA cm⁻²) and high cost.^1,10,11 Various efforts have been focused on improving its efficiency at high operating current density.^3,12–14 With the development of cell structure, electrode material and structure, and electrolyte additives, the polarization is dramatically reduced and the highest operating current density reaches 80 mA cm⁻² with high energy efficiency (80%),^3,14 but this still cannot fulfill the requirement for rapid charge and discharge. The exchange current density for zinc deposition and dissolution is found to be as high as 370 mA cm⁻² in alkaline solutions,¹⁵ which is considerably higher than the operating current density of ZNBs. Such ZNBs could be theoretically operated at several hundred milliamperes per square centimeter. However, to date, no study has been focused on the battery performance, potential problems and the critical obstacle for the improvement of battery performance at such high operating current density.

Here, the performance of a ZNB at high current density (up to 300 mA cm⁻²) has been investigated, and the overpotentials of the positive and negative electrodes are quantitatively analyzed to search for a critical obstacle. The morphologies of the deposited zinc and potential problems at high current densities have been analyzed. Based on these findings, we point out the remaining issues and struggling directions enabling high power density ZNBs.

2. Experimental section

2.1 Chemicals

Electrolyte was an aqueous solution of alkaline potassium zincate made of 0.4 mol dm⁻³ zinc oxide and 8 mol dm⁻³ potassium hydroxide. All the chemicals were dissolved in deionized water without further treatment.

2.2 Electrochemical measurements

The schematic diagram of a single cell structure is shown in Fig. 1. The cell employed nickel hydroxide electrode (30 × 30 × 0.7 mm, Jiangsu Highstar Battery Manufacturing, China) with a capacity of 25 mA h cm⁻² as the positive electrode, nickel foam (30 × 30 × 2 mm, 420 g m⁻², 110 PPI, Changsha Lyrun Material Co., Ltd., China) as the negative electrode and 8 mol dm⁻³ potassium hydroxide solution mixed with 0.4 mol dm⁻³ zinc oxide as the electrolyte. The distance between the positive electrode and negative electrode was 5 mm. The electrolyte was circulated through the cell by a pump with a rate of 19 cm s⁻¹. The cell was charged to 80% capacity based on the capacity of the positive electrode and discharged to 0.8 V under the same constant current density from 40 mA cm⁻² to 300 mA cm⁻² demonstrated by Arbin BT-2000. The positive and negative potentials were measured with an Hg/HgO reference electrode (0.098 V vs. SHE) installed in the middle of the gap between the two electrodes. All the experiments were conducted at room temperature.

Fig. 1 The schematic diagram of a single cell structure.

2.3 SEM characterization

Negative electrodes were washed with deionized water and dried in a vacuum desiccator after charging to the capacity of 10 mA h cm⁻² at various current densities (40 mA cm⁻² to 300 mA cm⁻²). The morphologies of the deposited zinc were surveyed by scanning electron microscopy (SEM, JEOL JSM-6000, JEOL Ltd., Japan).

3. Results and discussion

The charge–discharge voltage curve is an important parameter that can evaluate charge voltage and discharge voltage at high current density. The charge voltage increases and the discharge voltage declines with the current density increasing from 40 mA cm⁻² to 300 mA cm⁻² (Fig. 2). Two critical indexes of battery application are the mid-output voltage (voltage value at the midpoint of the discharge period) and power density. The former decreases from 1.64 V to 1.32 V due to large polarization, and the latter improves from 65.7 mW cm⁻² to 395.3 mW cm⁻² with a slope of 1.26 mW mA⁻¹ from 40 mA cm⁻² to 300 mA cm⁻² (Fig. 3). At 200 mA cm⁻² for instance, the mid-output voltage (1.40 V) is higher than that of the mature vanadium–vanadium system (about 1.27 V)¹⁶ and the recently proposed metal free quinine–bromine system (about 0.64 V),¹⁷ which are the only two systems that can operate at such high current density. As a result, the power density is the highest in flow battery families.

Fig. 2 Charge–discharge voltage curves at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

Fig. 3 Mid-output voltage and power density at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

The coulombic efficiency (CE), voltage efficiency (VE) and energy efficiency (EE) are three important indexes of battery performance, which are defined by the ratios of discharge coulomb to charge coulomb, mean discharge voltage to mean charge voltage and discharge energy to charge energy, respectively (Fig. 4). There is an abnormal phenomenon that CE is about 98.2% at 80 mA cm⁻² and 120 mA cm⁻², and decreases with the decreasing current density to low current density or increasing to high current density. This is because of the non-uniform distribution of potential on the negative electrode, which makes the less utilized electrode surface available for the main reaction and provides more surface area for hydrogen evolution at a low current density. However, at high current density, more oxygen is produced on the positive electrode and hydrogen is produced on the negative electrode, which deteriorate the CE. As a result, the optimum operating current density range to achieve high coulombic efficiency is between 80 mA cm⁻² and 120 mA cm⁻² under current operating conditions. However, VE decreases from 86.4% to 58.3% with the current density increasing from 40 mA cm⁻² to 300 mA cm⁻² owing to its large polarization. As a result, EE decreases from 82.6% to 50.7%. This is the first report showing the trends in efficiencies of ZNBs at ultrahigh current density up to 300 mA cm⁻².

Fig. 4 Coulombic efficiency (CE), voltage efficiency (VE) and energy efficiency (EE) at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

The potentials of positive electrodes and negative electrodes are presented in Fig. 5 and 6. The potential of the positive electrode increases upon charging and decreases upon discharging with increasing current density, while the potential of the negative electrode decreases upon charging and increases upon discharging due to large polarization (sluggish electrochemical reactions and large ohmic resistance). There is an interesting observation that negative electrode potential suddenly drops at the end of the discharging process at a low current density (40 mA cm⁻² and 80 mA cm⁻²), which occurs on the positive electrode at high current density (120 mA cm⁻² to 300 mA cm⁻²). This indicates that the negative electrode presents serious concentration polarization at the end of the discharging process at low current density (40 mA cm⁻² and 80 mA cm⁻²), which shifts to the positive electrode at high current density from 120 mA cm⁻² to 300 mA cm⁻². The zinc deposited on the negative electrode completely dissolved at low current density (40 mA cm⁻² and 80 mA cm⁻²), and the positive charged state of NiOOH may not completely discharge. This resulted from more electric charge consumed by side reaction (mainly hydrogen evolution) at the negative side than that at the positive side (mainly oxygen evolution). However, at a high current density (120 mA cm⁻² to 300 mA cm⁻²), zinc does not completely dissolve during the discharging process and thus accumulates on the negative electrode causing a short circuit upon repeated charge–discharge cycles, and deep discharge occurs on the positive electrode, which inevitably causes significant damnification of the cycle life of the solid electrode. This phenomenon stemmed from less electric charge consumed by the side reaction at the negative side (mainly hydrogen evolution) than that at the positive side (mainly oxygen evolution) at a high current density. As discussed above, the electric charge consumed by side reactions varies at both positive electrode and negative electrode at different operating current densities. Moreover, the asymmetry coulombic efficiencies of the positive electrode and negative electrode will cause the deterioration of battery stability, and it should be avoided as much as possible.

Fig. 5 Positive electrode potential at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

Fig. 6 Negative electrode potential at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

Different operating current density will result in different morphologies of deposited zinc on the negative electrode, as shown in Fig. 7. With the current density increasing, the morphologies undergo smooth (Fig. 7a), spongy (Fig. 7b–e) and dendrite (Fig. 7f) corresponding to the current densities of 40 mA cm⁻², 80 to 200 mA cm⁻², and 300 mA cm⁻², respectively. The smooth deposits are formed at a low current density when the electrode reaction is controlled by electrochemical reactions. Dendritic deposits are formed when the rate of the electrode reaction is controlled by the mass transport of zincate species to the electrode surface. Similarly, the spongy deposits are formed at moderate reaction rates when the electrode reaction is controlled by the mix of mass transport of active species and electrochemical reactions. Thus, the morphologies of deposited zinc, which transported from smooth, spongy to dendrite with increasing current density also have a strong relationship with the operating current density. At a high current density, the mass transport of the active species should be enhanced to depress the dendritic deposition of zinc.

Fig. 7 Morphologies of the deposited zinc on the negative electrode at various current densities: (a) 40 mA cm⁻², (b) 80 mA cm⁻², (c) 120 mA cm⁻², (d) 160 mA cm⁻², (e) 200 mA cm⁻², (f) 300 mA cm⁻².

We calculated the overpotentials of the positive electrode and negative electrode at various current densities (Fig. 8, 40 mA cm⁻² to 300 mA cm⁻²) to find the critical obstacle for improving battery performance. The overpotential of the positive electrode increases from 107.7 mV to 330.1 mV and that of the negative electrode increases from 20.2 mV to 177.8 mV. The large overpotential of the positive electrode is the prominent contributor that causes many side reactions, high charge voltage and low output voltage, and hinders the improvement of battery performance, especially at high current density. Attempts should be made to format a high active positive electrode through the preparation of active materials and electrode structure to achieve high power density ZNBs.

Fig. 8 Mid-overpotential of positive electrode and negative electrode at various current densities (40 mA cm⁻² to 300 mA cm⁻²).

4. Conclusions

During the real charge–discharge cycling, there is an optimum operating current density range to achieve high coulombic efficiency at fixed conditions including the cell structure, flow rate of electrolyte, and concentration of electrolyte. Similarly, the morphologies of the deposited zinc, which transported from smooth, spongy to dendrite with increasing current density, also have a strong relationship with the operating current density. At a high current density, the mass transport of active species should be enhanced to depress the dendritic deposition of zinc. Furthermore, the unmatched coulombic efficiencies of the negative electrode and positive electrode may lead to decrease in the cycle life of ZNBs, which has not attracted enough attention. Moreover, the large overpotential of the positive electrode is the critical obstacle for the improvement of battery performance at high operating current density. The elucidation of the performance and potential problems of ZNBs provide valuable information for its development and modification.

Acknowledgements

This work is financial supported by the National Basic Research Program of China (973 Program no. 2010CB227204).

References

1. J. Cheng, L. Zhang, Y. Yang, Y. Wen, G. Cao and X. Wang, Electrochem. Commun., 2007, 9, 2639–2642.
2. Y. Ito, M. Nyce, R. Plivelich, M. Klein and S. Banerjee, J. Power Sources, 2011, 196, 6583–6587.
3. Y. Cheng, H. Zhang, Q. Lai, X. Li and D. Shi, Electrochim. Acta, 2013, 105, 618–621.
4. F. R. McLarnon and E. J. Cairns, J. Electrochem. Soc., 1991, 138, 645–664.
5. C. Xu, B. Li, H. Du and F. Kang, Angew. Chem., Int. Ed., 2012, 51, 933–935.
6. A. K. Shukla, S. Venugopalan and B. Hariprakash, J. Power Sources, 2001, 100, 125–148.
7. M. Oshitani, T. Takayama, K. Takashima and S. Tsuji, J. Appl. Electrochem., 1986, 16, 403–412.
8. Y. Wang, D. Cao, G. Wang, S. Wang, J. Wen and J. Yin, Electrochim. Acta, 2011, 56, 8285–8290.
9. V. Viswanathan, A. Crawford, D. Stephenson, S. Kim, W. Wang, B. Li, G. Coffey, E. Thomsen, G. Graff, P. Balducci, M. Kintner-Meyer and V. Sprenkle, J. Power Sources, 2014, 247, 1040–1051.
10. Y. Ito, M. Nyce, R. Plivelich, M. Klein, D. Steingart and S. Banerjee, J. Power Sources, 2011, 196, 2340–2345.
11. Y. Cheng, H. Zhang, Q. Lai, X. Li, Q. Zheng, X. Xi and C. Ding, J. Power Sources, 2014, 249, 435–439.
12. Y. Wen, T. Wang, J. Cheng, J. Pan, G. Cao and Y. Yang, Electrochim. Acta, 2012, 59, 64–68.
13. J. Cheng, Y. Wen, G. Cao and Y. Yang, J. Power Sources, 2011, 196, 1589–1592.
14. Y. Cheng, H. Zhang, Q. Lai, X. Li, D. Shi and L. Zhang, J. Power Sources, 2013, 241, 196–202.
15. J. O. M. Bockris, Z. Nagy and A. Damjanovic, J. Electrochem. Soc., 1972, 119, 285–295.
16. Q. Zheng, F. Xing, X. Li, T. Liu, Q. Lai, G. Ning and H. Zhang, J. Power Sources, 2014, 266, 145–149.
17. B. Huskinson, M. P. Marshak, C. Suh, S. Er, M. R. Gerhardt, C. J. Galvin, X. Chen, A. Aspuru-Guzik, R. G. Gordon and M. J. Aziz, Nature, 2014, 505, 195–198.

---