Understanding Ammonium Transport in Bioelectrochemical Systems towards its Recovery

We report an integrated experimental and simulation study of ammonia recovery using microbial electrolysis cells (MECs). The transport of various species during the batch-mode operation of an MEC was examined experimentally and the results were used to validate the mathematical model for such an operation. It was found that, while the generated electrical current through the system tends to acidify (or basify) the anolyte (or catholyte), their effects are buffered by a cascade of chemical groups such as the NH3/NH4+ group, leading to relatively stable pH values in both anolyte and catholyte. The transport of NH4+ ions accounts for ~90% of the total current, thus quantitatively confirming that the NH4+ ions serve as effective proton shuttles during MEC operations. Analysis further indicated that, because of the Donnan equilibrium at cation exchange membrane-anolyte/catholyte interfaces, the Na+ ion in the anolyte actually facilitates the transport of NH4+ ions during the early stage of a batch cycle and they compete with the NH4+ ions weakly at later time. These insights, along with a new and simple method for predicting the strength of ammonia diffusion from the catholyte toward the anolyte, will help effective design and operation of bioeletrochemical system-based ammonia recovery systems.


Modeling of electrochemical and chemical reactions in the MEC
Because of bacteria activity, acetate (Ac -) is oxidized on the anode to release electron and proton: Through the external circuit, the released electrons are transferred to the cathode and react with the oxygen from the aeration: Since both equations are directly related to the electron transfer, the consumption/production of related species within the system can be modeled as a function of current, and they are referred to as Faradaic consumption/production in the main text.
The concentration of a species in the anolyte/catholyte/membrane can change due to its transport, Faradaic reactions (if applicable) and acid-base type chemical reactions. The effects of the acid-base type reactions are modeled as a sink/source term. Here, we illustrate the treatment of these acid-base type reactions using the evolution of NH 4 + /NH 3 concentration inside the anolyte as an example. For this couple, their reaction NH H ⇆ NH and their transport flux collectively change their concentrations. In the anolyte, the evolution of their concentrations is governed by Assuming fast equilibrium for the acid-base reaction, the partition between the NH 4 + and NH 3 , where , is the chemical equilibrium constant for the reaction NH H ⇆ NH . Equation

S6
states that the ratio of the NH 4 + and NH 3 concentrations are directly related to the pH locally.
As shown in Fig. S1, for pH<7 NH 4 + counts more than 99.4% of the total ammonium/ammonia; for pH>9.5, NH 3 counts more than 64%. Similar mathematical treatment is applied to all other acid-base type reactions. A full list of the acid-base reactions and their equilibrium constant is provided in Table S2.   Table S1. Using conservation law, Equ. 7 is obtained as (reproduced here for convenience)   OH -4.047×10 -4 6.024×10 -5 * The initial species concentration in catholyte is mostly practically 0 in the validation simulation, the same as in the experiment. In the parametric study, the initial species concentration is the same unless otherwise denoted.

Ionic competition
As explained in the main text, due to the opposite direction of their diffusion and migration through the CEM, Na + carries little charge across the CEM and the ion competition effect between Na + and NH 4 + is moderate. Here we assess the ion competition effect when the Na + ion concentration inside the anolyte is even higher than that of the NH 4 + ions. As shown in Fig. S2, for higher initial concentration of Na + ions, the backward transport of Na + ions from the catholyte into the anolyte and associated positive effect on NH 4 + ions removal from the anolyte during the early stage of operation decreases moderately. The competition of Na + ions for transport across CEM sets in earlier than that under lower Na + ion concentration in the anolyte.
For lower initial concentration of Na + ions, the opposite occurs. To further assess the ion competition effect, we examined the NH 4 + concentration in anolyte during MEC operation with different initial Na + concentration in the anolyte while keeping all other operating conditions the same. Figure S3 shows that, as the initial Na + concentration in the anolyte increases (decreases), less (more) NH 4 + ions are transported out of the anolyte due to stronger ion competition. Most of difference occurs during the early stage of operation, and the difference in the later stage is small. Overall, the transport of NH 4 + ions out of the anolyte is affected only moderately by the initial Na + ion concentration in the anolyte, suggesting that the ion competition effect is moderate in the system studied here.