Increasing the Cycle Life of Zinc Metal Anodes and Nickel-Zinc Cells Using Flow-Through Alkaline Electrolytes

From classical transport theory, conformal electrodepositions of many metals can only be sustained below a diffusion-limited critical value called the limiting current density (${i}_{L}$). ³ Current densities above ${i}_{L}$ can trigger electroconvection hydrodynamic instabilities and dendrite formation. In (i) LSV tests, the limiting currents are experimentally determined by LSV tests in stagnant electrolyte. Figure s1 shows ${i}_{L}$ = 25.5, 43.1, 86.5 mA cm⁻² for 0.2, 0.4, 0.6 M ZnO electrolyte, respectively.

The dimensionless Peclet number is:

$$\begin{eqnarray}&&Pe={Re}\cdot {Sc}\end{eqnarray} \tag{ 1 }$$

where $\mathrm{Re}\,=\,{du}/v$ is the Reynolds number and ${Sc}=\nu /D$ is the Schmidt number. Here, $d$ is a characteristic length scale, $u$ is the flow velocity, $\nu$ is the kinematic viscosity, and $D$ is the diffusivity of zincate ions. For the mesh anode, the characteristic length is the wire diameter, ^47–49 d = 30 μm. At Pe = 1, the convective transport equals the diffusive transport. While the electrochemical reactions listed in the Introduction section suggest that a hydroxide ion concentration gradient would develop between the two electrode-electrolyte interfaces, zincate ion concentration gradients may also develop in a small region close to the electrode-electrolyte interface. Thus, Zn plating/stripping is expected to be affected by the diffusivity of zincate and hydroxide ion. Here, zincate ion diffusivity is chosen for Pe calculations as the low concentration of zincate ions suggests that the zincate ion diffusion would be the determining factor that governs the transition from under-limiting to over-limiting regime. The zincate diffusivity coefficient is independent of the zincate ion concentration over the range of 0.02–0.4 M. ⁵⁰ Here, we choose D to be $3.8\times {10}^{-10}$ m² s⁻¹. The critical flow rate ⁴² corresponding to Pe = 1 is 0.0132 mm s⁻¹.

Test (ii) Electrodeposition tests are conducted to investigate the influence of flow on the electrodeposition morphology. To fully dissolve concentrated 0.6 M ZnO, we choose 10 M KOH electrolyte. Figure 2 shows the Zn morphology captured by SEM after 3 min of electrodeposition at the diffusion-limit. In stagnant electrolytes, the deposited zinc is dendritic at 0.2, 0.4, and 0.6 M ZnO concentrations. Under continuous ion consumption at the zinc electrode surface, the zincate ion concentration gradient increases with time, eventually causing ion depletion. The ion depletion further induces electroconvection at the electrode surface that accelerates the dendrite formation. ³ Figure 2b shows that fresh zincate ions carried by flowing electrolyte with small flow rate (Pe = 1, v = 0.0132 mm s⁻¹) can reduce the dendrite formation in 0.2 M ZnO electrolyte. Increasing the flow rate to Pe = 10 produces dendrite-free electrodeposition by bringing sufficient zincate ions to the anode surface as shown in Fig. 2c. Similar phenomena are observed in 0.4 and 0.6 M ZnO electrolytes. During the galvanostatic electrodeposition process, we observe severe HER in stagnant electrolytes. Figure 3 shows that gas generation increases cell polarization and affects the voltage response due to continuous bubble formation and detachment at the electrode surface in static electrolyte. HER can also accelerate the zinc anode passivation, local pH in stability and dendrite formation, causing capacity loss. ¹⁴ In Fig. 3, gas evolution is reduced at Pe = 1 and fully suppressed at Pe = 10. The flow also decreases the cell overpotential in these three electrolytes as shown in Figs. 3 and s2.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

To investigate the reversibility and efficiency of zinc plating/stripping at different flow conditions and applied currents, continuous plating/stripping is conducted in Test (iii) Asymmetric cell tests: (iii.1) ${i}_{L}$ for 6.5 mAh cm⁻², (iii.2) ${i}_{L}$ for 13 mAh cm⁻², and (iii.3) ${2i}_{L}$ for 2.6 mAh cm⁻². The two electrodes are a Zn-plated Cu wire mesh and a Cu wire mesh. The cell capacity is affected by the initial capacity which is determined by the amount of pre-deposited Zn on the Cu electrode. Here, we choose 6 M KOH because it is sufficient to fully dissolve 0.2 M ZnO. All tests charge/discharge to $\pm$ 1.3 V cutoff voltages to ensure that Zn is fully stripped from the Cu anode. Figure (a) shows the voltage profiles under stagnant and Pe = 10 flow conditions for test (iii.1). Internal shorts are not observed in test (iii.1), either without flow or with flow at Pe = 10. Flowing electrolyte, however, affects the capacity and CE of these two cells as shown in Fig. 4b. Capacity decreases with cycle numbers due to active Zn loss. In stagnant electrolyte, the capacity quickly drops to 50% of its initial capacity in the 9th cycle and is approximately zero after 20 cycles. At Pe = 10, the cell maintains at least 50% initial capacity until the 53rd cycle, approximately 6 times the cycle life of the cell in stagnant electrolyte.

Zoom In Zoom Out Reset image size

Test (iii.2) doubles the initial capacity compared to test (iii.1) to investigate its impact on asymmetric cycle reversibility. Figure 4c shows that the asymmetric cell in stagnant electrolyte experiences an internal short in 20 cycles, while the flow-assisted cell shows no sign of an internal short. In Fig. s3b, the stagnant cell loses 50% of its initial capacity in the 5th cycle. At a relatively small flow rate (Pe = 10), the cell maintains 50% of its initial capacity for 29 cycles at an average CE of 98.6%, almost 6 times longer. The results show that the lifespan decreases with increased electrodeposition capacity. Compared with the stagnant electrolyte, the flow-through electrolyte can prolong the cell lifespan by avoiding internal short due to dendrite formation.

In test (iii.3), we apply ${2i}_{L}$ (52 mA cm⁻²) to the cell for 3 min to investigate cycling in the over-limiting regime. In Fig. 4d, the asymmetric cell under no flow condition loses capacity after 23 cycles. The CE and capacity are shown in Fig. s4. Severe HER is observed in stagnant electrolyte and the cell loses 50% of its initial capacity by the 6th cycle (see Fig. s4). At a flow rate of Pe = 20, the cell cycles for over 90 cycles at 50% of its initial capacity.

Previous investigations prove that flow-through electrolyte can improve the efficiency and deposition morphology in Zn-Cu cells. Flow-through electrolyte is then applied to Ni-Zn cells.

Figure s5 compares no flow with three flow directions at Pe = 1 in a Ni-Zn cell. The flow direction (forward, backward, oscillating) does not affect voltage response during charge/discharge. Fresh zincate ions are brought to the anode surface regardless of flow direction. All flow directions result in improved battery impedance compared to the no flow condition.

A Ni-Zn cell with 0.2 M ZnO in 6 M KOH is cycled with a constant charging current density of ${\boldsymbol{2}}{{\boldsymbol{i}}}_{{\boldsymbol{L}}}$ (52 mA cm⁻²) to 6 mAh cm⁻², followed by a constant current discharge at ${{\boldsymbol{i}}}_{{\boldsymbol{L}}}$ (26 mA cm⁻²) to 1.4 V, corresponding to approximately 100% DoD where the deposited zinc is fully stripped from the Cu anode. In the cycling tests, we disassemble the commercial Ni-Zn battery to obtain the Ni(OH)₂/NiO(OH) electrode. Based on the data sheet, the utilized commercial Ni(OH)₂/NiO(OH) electrode has a maximum capacity of approximately 375 mWh (1.6 V). Figures 5a and 5b show the CE and voltage profiles without and with (Pe = 77) flow. The CE in stagnant electrolyte drops to 50% by the 34th cycle, and it only retains 31% of its initial capacity after 77 cycles due to dendrite formation and side reactions. A high flow rate of Pe = 77 (1 mm s⁻¹) significantly suppresses dendrite formation and side reactions. The cell exhibits over 95% CE for 297 cycles. The increased polarization at Pe = 77 after the 34th cycle is most likely due to the poor cycle performance of the Ni(OH)₂/NiO(OH) electrode at high charging rates. ⁵¹

Zoom In Zoom Out Reset image size

To focus on the Zn anode performance and reduce the influence of the cathode rate limits, the Ni(OH)₂/NiO(OH) electrode is oversized. The utilized commercial Ni(OH)₂/NiO(OH) electrode has a maximum capacity of approximately 750 mWh (1.6 V). The applied charge/discharge current density is 35 mA cm⁻² (1.4 ${{\boldsymbol{i}}}_{{\boldsymbol{L}}}$) based on anode surface area. The battery is charged/discharged at approximately 5 C rate, with each cycle charging to 7 mAh cm⁻² and discharging to a cutoff voltage of 1.4 V. The electrolyte is 6 M KOH with 0.2 M ZnO.

Figures 6a and 6b show CE and voltage, respectively, at Pe = 0 and 20 (0.26 mm s⁻¹). In stagnant electrolyte, the Ni-Zn battery experiences an internal short after 88 cycles and only lasts for 30 h, with a CE drop to 43% by the 32^nd cycle. However, at Pe = 20, the battery exhibits significantly improved reversibility with over 95% CE after 1500 cycles (600 h).

Zoom In Zoom Out Reset image size

Ni-Zn batteries typically experience capacity decay primarily due to Zn loss and side reactions. In our tests, the cathode is oversized with a maximum initial capacity of approximately 750 mWh (1.6 V). This allows us to focus on evaluating the Zn anode performance and mitigate the influence of cathode rate limits. The selected charge capacity of 7 mAh cm⁻² represents less than 10% of the oversize cathode capacity. Also, the cell is flooded with electrolyte with sufficient ZnO to compensate for capacity loss resulting from Zn stripping.

To achieve high capacity, applying flowing electrolyte to batteries consisting of multiple layers of electrodes with a compact, flat-stack configuration is the focus of future work. Reduced porosity resulting from Zn deposition can impact various factors such as back pressure, flow path, and pumping power loss. Future work will explore the consequences of reduced porosity on battery performance and fluid mechanics.