Youneng Tang

Interactions between Perchlorate and Nitrate Reductions in the Biofilm of a Hydrogen-Based Membrane Biofilm Reactor

We studied the microbial functional and structural interactions between nitrate (NO(3)(-)) and perchlorate (ClO(4)(-)) reductions in the hydrogen (H(2))-based membrane biofilm reactor (MBfR). When H(2) was not limiting, ClO(4)(-) and NO(3)(-) reductions were complete, and the MBfRs biofilm was composed mainly of bacteria from the ε- and β-proteobacteria classes, with autotrophic genera Sulfuricurvum, Hydrogenophaga, and Dechloromonas dominating the biofilm. Based on functional-gene and pyrosequencing assays, Dechloromonas played the most important role in ClO(4)(-) reduction, while Sulfuricurvum and Hydrogenophaga were responsible for NO(3)(-) reduction. When H(2) delivery was insufficient to completely reduce both electron acceptors, NO(3)(-) reduction out-competed ClO(4)(-) reduction for electrons from H(2), and mixotrophs become important in the MBfR biofilm. β-Proteobacteria became the dominant class, and Azonexus replaced Sulfuricurvum as a main genus. The changes suggest that facultative, NO(3)(-)-reducing bacteria had advantages over strict autotrophs when H(2) was limiting, because organic microbial products became important electron donors when H(2) was severely limiting.

Nitrate shaped the selenate-reducing microbial community in a hydrogen-based biofilm reactor.

To study the effect of nitrate (NO3(-)) on selenate (SeO4(2-)) reduction, we tested a H2-based biofilm with a range of influent NO3(-) loadings. When SeO4(2-) was the only electron acceptor (stage 1), 40% of the influent SeO4(2-) was reduced to insoluble elemental selenium (Se(0)). SeO4(2-) reduction was dramatically inhibited when NO3(-) was added at a surface loading larger than 1.14 g of N m(-2) day(-1), when H2 delivery became limiting and only 80% of the input NO3(-) was reduced (stage 2). In stage 3, when NO3(-) was again removed from the influent, SeO4(2-) reduction was re-established and increased to 60% conversion to Se(0). SeO4(2-) reduction remained stable at 60% in stages 4 and 5, when the NO3(-) surface loading was re-introduced at ≤ 0.53 g of N m(-2) day(-1), allowing for complete NO3(-) reduction. The selenate-reducing microbial community was significantly reshaped by the high NO3(-) surface loading in stage 2, and it remained stable through stages 3-5. In particular, the abundance of α-Proteobacteria decreased from 30% in stage 1 to less than 10% of total bacteria in stage 2. β-Proteobacteria, which represented about 55% of total bacteria in the biofilm in stage 1, increased to more than 90% of phylotypes in stage 2. Hydrogenophaga, an autotrophic denitrifier, was positively correlated with NO3(-) flux. Thus, introducing a NO3(-) loading high enough to cause H2 limitation and suppress SeO4(2-) reduction had a long-lasting effect on the microbial community structure, which was confirmed by principal coordinate analysis, although SeO4(2-) reduction remained intact.

Using a two-stage hydrogen-based membrane biofilm reactor (MBfR) to achieve complete perchlorate reduction in the presence of nitrate and sulfate.

We evaluated a strategy for achieving complete reduction of perchlorate (ClO(4)(-)) in the presence of much higher concentrations of sulfate (SO(4)(2-)) and nitrate (NO(3)(-)) in a hydrogen-based membrane biofilm reactor (MBfR). Full ClO(4)(-) reduction was achieved by using a two-stage MBfR with controlled NO(3)(-) surface loadings to each stage. With an equivalent NO(3)(-) surface loading larger than 0.65 ± 0.04 g N/m(2)-day, the lead MBfR removed about 87 ± 4% of NO(3)(-) and 30 ± 8% of ClO(4)(-). This decreased the equivalent surface loading of NO(3)(-) to 0.34 ± 0.04-0.53 ± 0.03 g N/m(2)-day for the lag MBfR, in which ClO(4)(-) was reduced to nondetectable. SO(4)(2-) reduction was eliminated without compromising full ClO(4)(-) reduction using a higher flow rate that gave an equivalent NO(3)(-) surface loading of 0.94 ± 0.05 g N/m(2)-day in the lead MBfR and 0.53 ± 0.03 g N/m(2)-day in the lag MBfR. Results from qPCR and pyrosequencing showed that the lead and lag MBfRs had distinctly different microbial communities when SO(4)(2-) reduction took place. Denitrifying bacteria (DB), quantified using the nirS and nirK genes, dominated the biofilm in the lead MBfR, but perchlorate-reducing bacteria (PRB), quantified using the pcrA gene, became more important in the lag MBfR. The facultative anaerobic bacteria Dechloromonas, Rubrivivax, and Enterobacter were dominant genera in the lead MBfR, where their main function was to reduce NO(3)(-). With a small NO(3)(-) surface loading and full ClO(4)(-) reduction, the dominant genera shifted to ClO(4)(-)-reducing bacteria Sphaerotilus, Rhodocyclaceae, and Rhodobacter in the lag MBfR.

Effects of multiple electron acceptors on microbial interactions in a hydrogen-based biofilm

To investigate interactions among multiple electron acceptors in a H2-fed biofilm, we operated a membrane biofilm reactor with H2-delivery capacity sufficient to reduce all acceptors. ClO4(-) and O2 were input electron acceptors in all stages at surface loadings of 0.08 ± 0.006 g/m(2)-d (1.0 ± 0.7 e(-) meq/m(2)-d) for ClO4(-) and 0.51 g/m(2)-d (76 e(-) meq/m(2)-d) for O2. SO4(2-) was added in Stage 2 at 3.77 ± 0.39 g/m(2)-d (331 ± 34 e(-) meq/m(2)-d), and NO3(-) was further added in Stage 3 at 0.72 ± 0.03 g N/m(2)-d (312 ± 13 e(-) meq/m(2)-d). At steady state for each stage, ClO4(-), O2, and NO3(-) (when present in the influent) were completely reduced; measured SO4(2-) reduction decreased from 78 ± 4% in Stage 2 to 59 ± 4% in Stage 3, when NO3(-) was present. While perchlorate-reducing bacteria (PRB), assayed by qPCR targeting the pcrA gene, remained stable throughout, sulfate-reducing bacteria (SRB), assayed by the dsrA gene, increased almost 3 orders of magnitude when significant SO4(2-) reduction occurred in stage 2. The abundance of denitrifying bacteria (DB), assayed by the nirK and nirS genes, increased in Stage 3, while SRB remained at high numbers, but did not increase. Based on pyrosequencing analyses, β-Proteobacteria dominated in Stage 1, but ε-Proteobacteria became more important in Stages 2 and 3, when the input of multiple electron acceptors favored genera with broader electron-accepting capabilities. Sulfuricurvum (a sulfur oxidizer and NO3(-) reducer) and Desulfovibrio (a SO4(2-) reducer) become dominant in Stage 3, suggesting redox cycling of sulfur in the biofilm.

A Steady-State Biofilm Model for Simultaneous Reduction of Nitrate and Perchlorate, Part 2: Parameter Optimization and Results and Discussion

Part 1 of this work developed a steady-state, multispecies biofilm model for simultaneous reduction of nitrate and perchlorate in the H(2)-based membrane biofilm reactor (MBfR) and presented a novel method to solve it. In Part 2, the half-maximum-rate concentrations and inhibition coefficients of nitrate and perchlorate are optimized by fitting data from experiments with different combinations of influent nitrate and perchlorate concentrations. The model with optimized parameters is used to quantitatively and systematically explain how three important operating conditions (nitrate loading, perchlorate loading, and H(2) pressure) affect nitrate and perchlorate reduction and biomass distribution in these reducing biofilms. Perchlorate reduction and accumulation of perchlorate-reducing bacteria (PRB) in the biofilm are affected by four promotion or inhibition mechanisms: simultaneous use of nitrate and perchlorate by PRB and competition for H(2), the same resources in PRB, and space in a biofilm. For the hydrogen pressure evaluated experimentally, a low nitrate loading (<0.1 g N/m(2)-d) slightly promotes perchlorate removal, because of the beneficial effect from PRB using both acceptors. However, a nitrate loading of >0.6 g N/m(2)-d begins to inhibit perchlorate removal, as the competition effects become dominant.

A pH-control model for heterotrophic and hydrogen-based autotrophic denitrification

This work presents a model to predict the alkalinity, pH, and Langelier Saturation Index (LSI) in heterotrophic and H(2)-based autotrophic denitrification systems. The model can also be used to estimate the amount of acid, e.g. HCl, added to the influent (method 1) or the pH set point in the reactor (method 2: pH can be maintained stable by CO(2)-sparge using a pH-control loop) to prevent the pH from exceeding the optimal range for denitrification and to prevent precipitation from occurring. The model was tested with two pilot plants carrying out denitrification of groundwater with high hardness: a heterotrophic system using ethanol as the electron donor and an H(2)-based autotrophic system. The measured alkalinity, pH, and LSI were consistent with the model for both systems. This work also quantifies: (1) how the alkalinity and pH in Stage-1 significantly differ from those in Stage-2; (2) how the pH and LSI differ significantly in the two denitrification systems while the alkalinity increase is about the same; and (3) why CO(2) addition is the preferred method for autotrophic system, while HCl addition is the preferred method for the heterotrophic system.

A steady-state biofilm model for simultaneous reduction of nitrate and perchlorate, part 1: model development and numerical solution.

A multispecies biofilm model is developed for simultaneous reduction of nitrate and perchlorate in the H(2)-based membrane biofilm reactor. The one-dimension model includes dual-substrate Monod kinetics for a steady-state biofilm with five solid and five dissolved components. The solid components are autotrophic denitrifying bacteria, autotrophic perchlorate-reducing bacteria, heterotrophic bacteria, inert biomass, and extracellular polymeric substances (EPS). The dissolved components are nitrate, perchlorate, hydrogen (H(2)), substrate-utilization-associated products, and biomass-associated products (BAP). The model explicitly considers four mechanisms involved in how three important operating conditions (H(2) pressure, nitrate loading, and perchlorate loading) affect nitrate and perchlorate removals: (1) competition for H(2), (2) promotion of PRB growth due to having two electron acceptors (nitrate and perchlorate), (3) competition between nitrate and perchlorate reduction for the same resources in the PRB: electrons and possibly reductase enzymes, and (4) competition for space in the biofilm. Two other special features are having H(2) delivered from the membrane substratum and solving directly for steady state using a novel three-step approach: finite-difference for approximating partial differential and/or integral equations, Newton-Raphson for solving nonlinear equations, and an iterative scheme to obtain the steady-state biofilm thickness. An example result illustrates the models features.

A biofilm model to understand the onset of sulfate reduction in denitrifying membrane biofilm reactors

This work presents a multispecies biofilm model that describes the co‐existence of nitrate‐ and sulfate‐reducing bacteria in the H2‐based membrane biofilm reactor (MBfR). The new model adapts the framework of a biofilm model for simultaneous nitrate and perchlorate removal by considering the unique metabolic and physiological characteristics of autotrophic sulfate‐reducing bacteria that use H2 as their electron donor. To evaluate the model, the simulated effluent H2, UAP (substrate‐utilization‐associated products), and BAP (biomass‐associated products) concentrations are compared to experimental results, and the simulated biomass distributions are compared to real‐time quantitative polymerase chain reaction (qPCR) data in the experiments for parameter optimization. Model outputs and experimental results match for all major trends and explain when sulfate reduction does or does not occur in parallel with denitrification. The onset of sulfate reduction occurs only when the nitrate concentration at the fibers outer surface is low enough so that the growth rate of the denitrifying bacteria is equal to that of the sulfate‐reducing bacteria. An example shows how to use the model to design an MBfR that achieves satisfactory nitrate reduction, but suppresses sulfate reduction. Biotechnol. Bioeng. 2013; 110: 763–772.

An improved cellular automaton method to model multispecies biofilms

Biomass-spreading rules used in previous cellular automaton methods to simulate multispecies biofilm introduced extensive mixing between different biomass species or resulted in spatially discontinuous biomass concentration and distribution; this caused results based on the cellular automaton methods to deviate from experimental results and those from the more computationally intensive continuous method. To overcome the problems, we propose new biomass-spreading rules in this work: Excess biomass spreads by pushing a line of grid cells that are on the shortest path from the source grid cell to the destination grid cell, and the fractions of different biomass species in the grid cells on the path change due to the spreading. To evaluate the new rules, three two-dimensional simulation examples are used to compare the biomass distribution computed using the continuous method and three cellular automaton methods, one based on the new rules and the other two based on rules presented in two previous studies. The relationship between the biomass species is syntrophic in one example and competitive in the other two examples. Simulation results generated using the cellular automaton method based on the new rules agree much better with the continuous method than do results using the other two cellular automaton methods. The new biomass-spreading rules are no more complex to implement than the existing rules.

Selenate and Nitrate Bioreductions Using Methane as the Electron Donor in a Membrane Biofilm Reactor

Selenate (SeO4(2-)) bioreduction is possible with oxidation of a range of organic or inorganic electron donors, but it never has been reported with methane gas (CH4) as the electron donor. In this study, we achieved complete SeO4(2-) bioreduction in a membrane biofilm reactor (MBfR) using CH4 as the sole added electron donor. The introduction of nitrate (NO3(-)) slightly inhibited SeO4(2-) reduction, but the two oxyanions were simultaneously reduced, even when the supply rate of CH4 was limited. The main SeO4(2-)-reduction product was nanospherical Se(0), which was identified by scanning electron microscopy coupled to energy dispersive X-ray analysis (SEM-EDS). Community analysis provided evidence for two mechanisms for SeO4(2-) bioreduction in the CH4-based MBfR: a single methanotrophic genus, such as Methylomonas, performed CH4 oxidation directly coupled to SeO4(2-) reduction, and a methanotroph oxidized CH4 to form organic metabolites that were electron donors for a synergistic SeO4(2-)-reducing bacterium.