Bipolar Polymer Membranes (BPMs) enable the creation of large, stable pH gradients by driving water dissociation (WD) at the cation/anion junction under reverse bias, a process central to electrodialysis, CO2 capture, and emerging acid–alkaline water electrolysis. Yet despite decades of study, the mechanism by which intense interfacial electric fields accelerate WD remains debated and is often modeled with ad hoc assumptions. In this study, we present a power dissipation model in which minority ions from water autoprotolysis act as carriers that continuously dissipate field-supplied power in the hydrated nanometric junction. This dissipative input increases the local probability of heterolytic O–H bond cleavage and analytically leads to a quadratic dependence of the dissociation rate constant on the field. Without adjustable parameters, the model reproduces the required orders of magnitude for the enhancement ratio kd(E)/kd(0), where kd(E) is the field-enhanced water dissociation rate constant and kd(0) is its zero-field value across typical BPM fields, and yields a quadratic current–voltage junction law. A proof-of-principle measurement on a commercial Fumasep® FBM bipolar membrane confirms the quadratic current–voltage trend, supporting a power-dissipation-driven water dissociation mechanism and providing a concise, falsifiable baseline for future studies.
Water dissociation (WD) at the internal junction of bipolar membranes (BPMs) is the key process enabling acid/base generation under reverse bias, yet the physical origin of its strong enhancement remains debated. In our previous work, we proposed a power-dissipation model in which WD is enhanced by an intense electric field through local power dissipation by autoprotolysis ions. Here, we experimentally validate this model using three commercial BPMs under acid/base and neutral salt conditions and extend it to high current density by incorporating finite water supply to the BPM junction. Under acid/base conditions, two BPMs showed strong field-dominated quadratic behavior, while a catalyst-containing BPM displayed a predominantly linear response, indicating that heterogeneous interfacial catalysis can mask the purely field-driven signature. In neutral , all three BPMs exhibited excellent quadratic fits, confirming that the predicted behavior is robust. The fitted prefactor varied with membrane type and electrolyte configuration, reflecting differences in hydration, junction thickness, transport behavior, and catalysis effect. The model was extended at higher current density to include finite diffusive water supply to the junction. The resulting saturation law links the intrinsic quadratic WD current to a water-transport-limited current , and predicts an inflection point at .