Aromatic compounds have always been of concern regarding their toxicity to living organisms, including microorganisms. With more anthropogenic activities (e.g. coal gasification), the need for feasible treatment of industrial effluents is highly prioritised. With the anaerobic degradation process being a competitive solution, these compounds’ toxic impact on the biomass is still of concern. These implications influence the stability of the degradation process; thus, there was a search for mechanisms to make the anaerobic degradation process more resilient. One potential mechanism is enhancing syntrophic collaboration between different species and its corresponding electron transfer. Syntrophic collaboration in an anaerobic environment can be conducted using intermediates (e.g. hydrogen) or direct electron transfer. Direct interspecies electron transfer (DIET) is reported to be more energy efficient and more thermodynamically favoured over other mechanisms that include mediators (hydrogen/ formic acid). Conductive and semi-conductive materials have been investigated to simulate this direct interspecies electron transfer mechanism (DIET), with various materials being researched, such as iron oxides, zero-valent metals, and even carbon-based materials.
This study investigated the impact of magnetite addition (as a DIET-stimulator) on p-cresol degradation, methane production and sludge characteristics, with a further interest in membrane fouling mitigation. This investigation was conducted with continuous flow reactors and batch reactors. The continuous configuration was based on an anaerobic membrane bioreactor (AnMBR) fed with a synthetic-coal gasification-like solution of phenol and p-cresol to investigate mainly the conversion rate of p-cresol and monitor the influence on the methane production, sludge characteristics, and membrane fouling. At the same time, batch experiments were conducted to investigate the acetoclastic methanogenic pathway and p-cresol degradation as a sole carbon source. The continuous experiment lasted for 143 days but was divided into two separate phases with two different magnetite dosages, starting with 40 mmol/L in phase (I), then replacing the sludge with acclimatised one (from the control) with the addition of the second dosage (20 mmol/L) in phase (II).
A Magnetite dosage of 40 mmol/L showed signs of biomass-suppressed conversion capacity compared to the control, by which the reactor conversion rate deteriorated by reaching 212 mgCOD/ gVSS/d (under a feed of 900 mgPh/L & 900 mgPcr/L). Phase (I) showed no significant differences in the methane production rate between control and magnetite reactors. On the other hand, the batch experiments fed with 1 gCOD/L acetate showed that the magnetite reactor had a lower acetoclastic methane production rate than the control. It was suggested that the 40 mmol/L magnetite dosage was suppressing the acetoclastic methanogens, which was further contributing to the lower conversion capacity observed in the AnMBR by the end of the phase. With the same methane being produced in control and magnetite reactors, it was also possible that either hydrogenotrophic methanogenesis or the DIET pathways were enhanced; however the absence of intermediates (e.g. VFAs) and the similarity of the COD balance supported the possibility of the latter one. During phase (II), the conversion rate of both reactors (control and magnetite) reached 74 mgPcr/ gVSS/d, approaching the highest conversion rates reported in the literature. While the acetoclastic methanogens showed no significant difference in the batch experiment, the magnetite-AnMBR’s methane production rate was 10%-28% higher. Furthermore, the methane yield with magnetite supplementation showed an average enhancement of 15%. In addition, the batch experiment also showed that this magnetite dosage reduced the p-cresol conversion rate by 87% compared to the control.
Both magnetite dosages (20 mmol/L & 40 mmol/L) showed a reduction in the protein and carbohydrate content of the soluble microbial products (SMP) and the extracellular polymeric substances (EPS). Magnetite had adversely impacted the loosely-bounded EPS (regarding protein and carbohydrates), whereas it was shown to be significant compared to the control. The EPS-LB showed an inverse relation with the particle size distribution (PSD), verifying that the higher increase in the particle size could be correlated with the EPS-LB reduction by the magnetite. On the other hand, the fouling rate of the membranes showed an insignificant difference between both reactors. This was suggested to be related to the incomplete formation of a mature cake layer under the influence of low operational flux. However, with the reduction of the SMP/EPS, it was suggested that the formed cake layer would be more porous and permeable. This would mean that the cake-fouling and its corresponding resistance would be expected to be lower. As the cake layer acts as a protective barrier for the membrane, its reduction would lead to a higher risk of irreversible pore-blocking by fine particles from the magnetite and the sludge.