Effects of grid alloy on the properties of positive-plate corrosion layers in lead/acid batteries. Implications for premature capacity loss under repetitive deep-discharge cycling service

Abstract

Premature capacity loss (PCL) has been demonstrated consistently by the deep-discharge cycling of three-plate lead/acid cells configured with an excess of electrolyte. A capacity loss of ∼2% per cycle was observed with cells based on tin-free, lead—calcium positive grids, under both constant-current and constant-voltage charging. The current that flows during constant-voltage charging decreases markedly within the first few cycles. This coincides with the establishment of an appreciable corrosion layer on the grid, and also with the onset of severe capacity loss. Significantly, there is no corresponding build-up of lead sulfate within the porous mass. With constant-current charging, a change in the overcharge factor, from 1.1 to 1.2 approximately doubles the rate of capacity loss. The corrosion products in lead—calcium plates exhibit a bi-layered structure: an outer corrosion layer of PbO 2 and an inner layer in which the composition approaches PbO. These materials are prone to fracture and separation, especially between the two layers. Cells based on lead—antimony positive grids also suffer PCL, but the rate of capacity loss (⊃1% per cycle is less than that observed for the lead—calcium analogues. The current under constant-voltage charging of lead—antimony cells also decreases during the first few charge/discharge cycles, yet, this effect is over-shadowed by an increase in current due to the effects of antimony migration. Increasing the level of overcharge under constant-current charging produces only a slight reduction in cycle-life. In regions close to the grid, PbO-like material is much less abundant than in lead—calcium plates. The corrosion products are composed mainly of PbO 2, and are more coherent under stress. Resistance at the grid/porous material interface, measured in situ, increases greatly during discharge for plates based on lead—calcium grids, but much less for the corresponding lead—antimony plates. This tends to emphasize the importance of the barrier-layer model of capacity loss, which is supported further by the observation of low-oxidation-state lead compounds in lead—calcium samples, but not in lead—antimony samples. The barrier-layer model cannot, however, provide a complete explanation for PCL: the corrosion layers of lead—antimony plates remain apparently conductive and coherent throughout cycle life, yet, these plates suffer appreciable capacity loss. At least part of the decrease in plate performance may, therefore be due to changes in the conductivity/integrity/activity of the bulk porous mass.