Microbial electrochemical cells (MxCs) are a novel technology that use anode-respiring bacteria (ARB) to bioremediate wastewaters and respire an electrical current, which can then be used directly to produce value-added products like hydrogen peroxide (H2O2). Ninety-five percent of the world’s H2O2 is currently produced using the anthraquinone process, whose production requires expensive and potentially carcinogenic catalysts and high amounts of electricity. However, the amount of H2O2 that can be produced from these microbial peroxide-producing cells (MPPCs) has not been thoroughly investigated. Predicting potential H2O2 production in MxCs is further complicated by a lack of mathematical models to predict performance utilizing complex waste streams like primary sludge (PS).

A reactor design methodology was developed for MPPCs to systematically optimize H2O2 production with minimal energy consumption. H2O2 stability was evaluated with different catholytes, membranes, and catalysts materials, and the findings used to design and operate long-term a dual-chamber, flat-plate MPPC using different catholytes, ferrochelating stabilizers, and hydraulic retention times (HRT). Up to 3.1 ± 0.37 g H2O2 L-1 was produced at a 4-h HRT in an MPPC with as little as 1.13 W-h g-1 H2O2 power input using NaCl catholytes. Attempts to improve H2O2 production by using weak acid buffers as catholytes or ferrochelating stabilizers failed for different reasons.

A non-steady-state mathematical model, MYAnode, was developed combinging existing wastewater treatment, anode biofilm, and chemical speciation models to predict MxC performance utilizing complex substrates. The model simulated the large-scale trends observed when operating an MPPC with PS substrate. At HRTs ≥ 12-d, the model demonstrated up to 20% Coulombic recovery. At these conditions, ARB required additional alkalinity production by ≥ 100 mgVSS/L of acetoclastic methanogens to prevent pH inhibition when little influent alkalinity is available. At lower HRTs, methanogens are unable to produce the alkalinity required to prevent ARB inhibition due to washout and rapid acidification of the system during fermentation. At ≥ 100 mgVSS/L of methanogens, increasing the diffusion layer thickness from 500 to 1000 μm improved Coulombic efficiency by 13.9%, while increasing particulate COD hydrolysis rates to 0.25/d only improved Coulombic efficiency by 3.9%.