Charge-Discharge Behavior of Zinc Electrodes in Alkaline Electrolytes with Water-Organic Mixed Solvents

Introduction

Zinc is an ideal negative electrode material for batteries with aqueous electrolytes, owing to its high reducing power of -1.26 V vs. SHE and high specific capacity of 820 mAh g^-1. However, zinc-based secondary batteries have not been established yet due to the insufficient rechargeability, as various deterioration modes of zinc anodes such as dendrite short-circuit, densification and shape changes lead to their limited cycle life and prevent the practical use as secondary batteries.

The origin of these deterioration modes are highly soluble nature of the discharge products (eq.1) tetrahydroxozincate anions (Zn(OH)₄^2-) in the alkaline electrolyte whose solubility is ca. 1 mol dm^-3 (M) in 8 M potassium hydroxide (KOH) aqueous solution for instance,  that are not stored in the vicinity of the electrode. The Zn(OH)₄^2-is finally decomposed and forms zinc oxide (ZnO) precipitation according to the subsequent chemical reaction (eq.2).

Zn + 4OH^- = Zn(OH)₄^2- + 2e^-                (1)

Zn(OH)₄^2-  = ZnO + 2OH^- + H₂O            (2)

In this paper, we focus on the water activity in equation 2 to reduce the zincate solubility. The reduction of water activity is expected to promote ZnO formation in the vicinity of electrode surface just after the zinc oxidation in equation 1 because equation 2 proceeds in the right side to lead to the soluble Zn(OH)₄^2-decomposition and ZnO precipitation. It is shown that the use of alkaline electrolyte containing organic solvents improves the cycle life at high zinc utilization rates. The effect of organic solvents on the zincate solubility including zincate supersaturation is examined and discussed.

Experimental

We prepared 4 M KOH alkaline electrolytes containing organic solvent by dissolving anhydrous alcohol such as propylene glycol (PG) and 8 M KOH aqueous solution. The alkaline electrolytes with PG-water mixed solvent was saturated with zincate species by mixing ZnO powder and removing residual solid by centrifugation.

We demonstrated charge-discharge tests with a three-electrode configuration. A working electrode (W.E.) was made of either planar zinc electrode deposited on copper plate or a porous ZnO electrode formed on the copper current collector. The W.E. electrode was sandwiched by two porous ZnO electrodes as counter electrodes inserted with non-woven separators. W.E. potential was monitored using a Hg/HgO reference electrode. The zinc cell was cycled with various utilization rate up to 75 % at various charge-discharge C-rate.

Results and Discussion

Figure 1 show the correlation between water pressure and zinc solubility in 4M KOH alkaline electrolyte containing the PG solvent. As shown in Fig.1 (a), the water pressure decrease with PG concentration monotonously, indicating the PG-water mixed solvent effectively suppress the water activity. Zincate solubility also decrease along with the suppression of the water activity. In addition to the thermostatic solubility of the zincate, the thermodynamic zincate supersaturation is also reduced to result in rapid precipitation of ZnO by using the PG-water mixed solvent (when compared with 4 M KOH).  

Using the alkaline electrolytes with PG-water mixed solvent, we evaluate charge-discharge cycle tests of zinc electrode. Figure 2 shows the cycle performance of zinc electrode with and without PG in 4 M KOH alkaline electrolyte at 75 % utilization of the theoretical capacity. In 4 M KOH without PG, the specific discharge capacity starts to decrease around 100^th cycle and achieve 1/3 of the initial capacity at 200^thcycle. On the other hand, with PG, the capacity keeps 75% utilization rate even after 900 cycles. After the cycle test, without PG, zinc was lost from the current collector, which was dissolved in alkaline electrolyte as supersaturated zincate species. On the contrary, with PG, zinc was left relatively uniform on copper substrate, owing to the suppression of zincate species dissolution in the PG-water mixed solvent. Other features such as suppression of dendrite formation and densification will be presented at the meeting.

Acknowledgment

This work was supported by RISING of NEDO.