This is a 2D cross-section of the same zinc-bromine flow cell, but the self-discharge mechanism is no longer a fitted formula — it is an actual Monte-Carlo diffusion simulation. The two half-reactions are:
Negative: Zn²⁺ + 2e⁻ ⇌ Zn(s) E° = -0.76 V
Positive: Br₂ + 2e⁻ ⇌ 2Br⁻ E° = +1.09 V
Cell: Zn²⁺ + 2Br⁻ ⇌ Zn(s) + Br₂ E°cell = 1.85 V
Charge/discharge still integrate current via Faraday's law (m = QM/(nF)) to grow the zinc deposit and move state of charge, and open-circuit voltage still follows the Nernst equation:
E = E°cell + (RT/nF)·ln( SOC / (1 - SOC) )
What's new here: bromine near the membrane-free separator is modeled as a pool of discrete tracer particles undergoing a symmetric random walk (a discretised 1D Fick's-law diffusion). Every particle that random-walks past the separator is a real crossing attempt — with probability equal to the complexing-agent ratio it is captured into the dense complex phase (sinks into the amber pool at the bottom-right); otherwise it escapes across unreacted, which is counted as one genuine self-discharge event that debits state of charge. The "Br₂ sequestration" and "Self-discharge" readouts are therefore measured from the simulated particle crossings, not computed from a closed-form coefficient.
- Complexing agent ratio — sets the per-attempt capture probability directly; watch the amber complex pool grow faster, and the self-discharge counter fall, as you raise it.
- Flow rate — widens the mass-transport boundary layer at low flow, raising overpotential (the gap between the running voltage and the idle Nernst curve), independent of the diffusion model.
- Real-world relevance — ZnBr flow batteries can be fully discharged to 0% SOC with no capacity fade, unlike Li-ion, because the "electrode" (plated Zn) is regrown fresh every cycle.