Elsevier

Electrochimica Acta

Volume 47, Issue 25, 25 September 2002, Pages 3995-4010
Electrochimica Acta

Modeling of an electrically rechargeable alkaline Zn–air battery

https://doi.org/10.1016/S0013-4686(02)00316-XGet rights and content

Abstract

The charging and discharge behavior of an electrically rechargeable alkaline Zn–air battery consisting of a porous Zn/ZnO electrode on the negative side and a porous O2 electrode on the positive side has been investigated. Galvanostatic experiments have been performed and a one-dimensional numerical model has been developed to analyze these experimental data. The cell voltages, the Zn electrode potentials versus Zn reference, and the O2 electrode potentials versus Zn reference calculated with this model are in fairly good agreement with the corresponding experimental values. The unwanted redistribution of Zn per cycle is predicted to decrease with increasing cycle number. The numerical model is expected to be useful when optimizing zinc cell designs for specific applications.

Introduction

Because of their high specific energy (>100 W h kg−1) and their inexpensive and environmentally benign materials, electrically rechargeable Zn–air batteries are promising power sources and energy storage devices [1], [2]. For their best possible utilization, it will be necessary to optimize their design with respect to known difficulties such as OH ion depletion occurring during discharge and shape change (redistribution of the active solid materials Zn and ZnO) occurring as a result of cycling.

Experimental studies of all critical parameters influencing the performance of a Zn–air battery are very time-consuming, hence numerical modeling can be very helpful in defining critical design parameters for the system such as, e.g. maximum Zn electrode thickness or minimum KOH concentration. A model of a rechargeable Zn–air battery has not been published so far, but models of similar systems have been presented and provide insight into the problems inherent in such systems. Choi et al. [3] developed a one-dimensional model of a rechargeable Zn–AgO battery in which they attributed shape change of the zinc electrode to convective flows chiefly caused by electro-osmotic forces. Sunu et al. [4] developed a one-dimensional model of a rechargeable Zn electrode while neglecting any finite kinetics at a counterelectrode. They concluded that a highly nonuniform reaction distribution will arise during high-current discharge which accentuates failures due to electrolyte depletion in the Zn electrode, resulting in a low discharge capacity and in shape change upon repeated cycling. Isaacson et al. [5] extended Sunu's model to investigate the movement of active material in directions both normal and parallel to the electrode surface in order to account for the effects of an electrolyte reservoir located on one side of the cell. Mao et al. [6] presented a one-dimensional model of a primary Zn–air battery to analyze its experimental discharge behavior.

In this work we develop a numerical model of an electrically rechargeable Zn–air battery allowing us to calculate from a galvanostatic experiment the response of cell voltage, Zn electrode potential (vs. Zn reference), O2 electrode potential (vs. Zn reference), and the potential and concentration profiles of the participating chemical species. We also present results from a galvanostatic three-cycle experiment, and from galvanostatic discharge experiments performed at high (24 mA cm−2), medium (12 mA cm−2), and low (1.28 mA cm−2) discharge current density. The model is then used to simulate these experimental data and to extend the calculation of the three-cycle experiment up to the end of cycle 40.

Section snippets

Experimental

The experimental setup included a twin-cell sandwich arrangement with a (negative) Zn electrode in the center and a (positive) O2 electrode on each side (Fig. 1). The anode, a porous Zn electrode, was prepared by a pasting method [7]. It consisted mainly of Zn and ZnO, but contained traces of PbO, cellulose, and PTFE. The PbO serves as inhibitor of corrosion reactions, while the cellulose fibers act as internal wicks to provide electrolyte channels, which keep the electrode wet and porous over

Model description

In this section we describe the physical background of our model and the corresponding set of equations used to calculate

  • i

    the cell voltage, the potential of the Zn electrode versus a Zn reference, and the potential of the O2 electrode versus a Zn reference, and

  • ii

    the concentration and potential profiles in the Zn electrode and in the separator

of a zinc–air battery during a galvanostatic experiment. The model battery consists of a single cell with one Zn and one O2 electrode, as shown in Fig. 4. The

Numerical solution

The fortran routine D03PDF commercially available from NAG (Numerical Algorithms Group) was used to solve the partial differential , . This routine integrates a system of partial differential equations in one space variable. The spatial discretization is performed using a Chebyshev C0 collocation method, and the method of lines is employed to reduce the PDE's to a system of ordinary differential equations. The resulting system is solved using a backward differentiation formula method. The

Calculations and results

The model described in Section 3 has been used to simulate the three-cycle and discharge experiments described in Section 2, and to extend the calculation of the three-cycle experiment up to the end of cycle 40. The values of the parameters used in these calculations are listed in Table 1, Table 2, Table 3. The kinetic parameters kstA, kfC, kstD, and kstF were optimized so as to get optimum agreement between calculated and experimental values of Ecell(t), EZn/Znref(t), and EO2/Znref(t). The

Calculated and observed values of Ecell(t), EZn/Znref(t), and EO2/Znref(t)

The calculated values of Ecell(t), EZn/Znref(t), and EO2/Znref (t) of the three-cycle experiment are in fairly good agreement with the experimental values (Fig. 2). The corresponding fit in the discharge experiments (Fig. 3) is good to about 80% of discharge, while beyond this point the calculated data reflect a higher state of discharge than the experiments. We attribute this deviation appearing beyond 80% of discharge to the decrease in electric conductivity of the Zn phase felt at low Zn

Conclusions

Galvanostatic three-cycle and discharge experiments have been performed, and a one-dimensional numerical model has been developed and was used to analyze the experiments. This model includes diffusion and migration of the participating dissolved species in the electrolyte of the porous Zn electrode and the separator, the chemical and electrochemical reactions occurring in the porous Zn electrode, as well as similar reactions occurring at the surface of the O2 electrode, which was assumed to be

List of symbols

Ageometric surface area of the electrode (cm2)
ck(x, t)concentration of the dissolved species k per unit of solution volume (mol cm−3)
cl*(x, t)concentration of the solid species l per unit of electrode volume (mol cm−3)
cO2concentration of O2 in the O2 electrode (mol cm−3)
crefreference concentration (1 mol dm−3)
Dkdiffusion coefficient of the dissolved species k (cm2 s−1)
dlgravimetic density of solid species l (cm3)
EO2Electric potential of the O2 electrode, taken as zero (or reference potential)

Acknowledgements

We thank Dr. K. Müller (Battelle Geneva) for a careful reading of the manuscript.

References (23)

  • J Goldstein et al.

    J. Power Sources

    (1999)
  • S Müller et al.

    Electrochim. Acta

    (1994)
  • E Yeager

    J. Mol. Catal.

    (1986)
  • E Yeager

    Electrochim. Acta

    (1984)
  • M.H Katz et al.

    J. Power Sources

    (1988)
  • R.E Davis et al.

    Electrochim. Acta

    (1967)
  • S Müller et al.

    Chimia

    (1995)
  • K.W Choi et al.

    J. Electrochem. Soc.

    (1976)
  • W.G Sunu et al.

    J. Electrochem. Soc.

    (1980)
  • M.J Isaacson et al.

    J. Electrochem. Soc.

    (1990)
  • Z Mao et al.

    J. Electrochem. Soc.

    (1992)
  • Cited by (133)

    View all citing articles on Scopus
    View full text