Removal of perchlorate from water using a biofilm magnetic ion exchange resin: feasibility and effects of dissolved oxygen, pH and competing ions

Abstract

A biofilm magnetic ion exchange (BMIEX) resin was obtained by mixing a magnetic ion exchange (MIEX) resin with perchlorate-acclimated cultures. Four-cycle batch tests using the BMIEX resin, virgin MIEX resin and cultures of perchlorate-reducing bacteria (PRB) were performed to investigate the feasibility and performance of the BMIEX resin in removing perchlorate (ClO₄⁻). This novel approach can achieve efficient simultaneous perchlorate adsorption and reduction through biofilm formation on the MIEX resin. Moreover, the effects of dissolved oxygen (DO), pH and common coexisting ions (NO₃⁻ and Cl⁻) on the BMIEX process were investigated to optimize perchlorate removal using BMIEX resin. Excellent perchlorate removal only occurred under anaerobic conditions, and a DO concentration of 2.07 mg L⁻¹ was sufficient to inhibit perchlorate removal by the BMIEX resin. A neutral pH range was optimal, and maximum perchlorate removal was obtained at pH 6.8. The presence of NO₃⁻ or Cl⁻ could adversely affect perchlorate reduction, and the inhibition of nitrate was substantially stronger than that of chloride. Increasing nitrate concentrations significantly inhibited the perchlorate removal efficiency of the BMIEX process. Community analysis of the biofilm revealed α-, β-, γ- and δ-Proteobacteria as the main PRB, accounting for approximately 21.04%, 12.44%, 24.33% and 0.71%, respectively.

---

1. Introduction

Perchlorate (ClO₄⁻), primarily used in the manufacturing of solid rockets, fireworks, and explosives, is a widespread inorganic contaminant in surface water and groundwater.^1–4 Perchlorate removal from contaminated waters has become a serious challenge to environmental engineers and scientists, primarily as a result of its high stability and mobility under typical environmental conditions.^5–7 Perchlorate, as an endocrine disruptor, competitively inhibits the uptake of iodide by the thyroid gland, resulting in the decreased synthesis and secretion of thyroid hormones.^8,9 In light of its potential negative impacts on human health, the U. S. EPA established a maximum of 15 μg L⁻¹ of ClO₄⁻ in drinking water sources in 2011.¹⁰

Among various treatment options, biological reduction and ion-exchange (IX) are prominent,^6,11,12 the latter of which has long been used to remove low levels of ClO₄^− 13 because of its simplicity, high capacity, short contact time and small footprint.^2,14 However, perchlorate is not destroyed in this process. Additionally, ion exchange has several shortcomings including high operational and regeneration costs, and complicated procedures involved in the treatment.⁴ Compared to IX, microorganisms can reduce perchlorate to non-toxic chloride (Cl⁻)¹¹ and biological processes also can be used to treat the high perchlorate concentration wastewater. However, biological technologies are not efficient at low concentrations because the degradation rates decreased obviously as the perchlorate concentration decreases; resulting in a long contact time and large size reactors. Moreover, when used in drinking water plants, perchlorate-reducing bacteria (PRB) need diverse carrier media, such as granular activated carbon (GAC), for biofilm colonization.¹⁵ To compensate for the shortcomings of these two technologies, methods utilizing ion exchange combined with biodegradation have been developed to remove ClO₄⁻ from aqueous solutions. Prior research has revealed several key findings in the direct bioregeneration of perchlorate-laden selective ion-exchange resins such as SBG1 and Purolite A530E/A532E.^{13,14,16–22} After exposing spent ion exchange resin to a high-strength PRB culture under anaerobic conditions in an offline system, the bioregenerated resin can then be effectively reused, rather than typically incinerated.^14,22 Furthermore, bioregeneration is more effective for macroporous than for gel-type resins, especially when macroporous resins have relatively small bead sizes with higher chloride concentrations.²⁰ However, this particular process required a long reaction time (8–20 days)²³ and a series of subsequent processing steps, including rinsing several times and disinfection to remove bio-fouling.^14,22

Magnetic ion exchange (MIEX) resin, marketed by the South Australian Water Corporation, has been found to adsorb perchlorate effectively.^24,25 The MIEX resin is a strong-base resin with iron oxide integrated into a macroporous, polyacrylic matrix, and is typically used with chloride as the exchangeable ion. However, the high-strength perchlorate brine generated during the regeneration process is a significant environmental problem. Accordingly, a novel bioregeneration concept, expressed as biofilm magnetic ion exchange (BMIEX), was developed that involves MIEX resin covered with a biofilm of PRB similar to biological activated carbon (BAC), combining the adsorption of perchlorate from water to MIEX resin and the subsequent bioreduction of perchlorate by the biofilm on the resin surface. This entire process is more efficient than either biodegradation or adsorption alone and concurrently regenerates the MIEX resin and biologically destroys perchlorate. The specific objectives of this research were to (1) evaluate longevity of the BMIEX resin technique and determine feasibility of perchlorate destruction rather than a simple enrichment by four cycles of batch experiments; (2) prove that the PRB in the BMIEX process play an important role in perchlorate destruction through the effect of dissolved oxygen (DO); (3) investigate the effects of different factors, such as the initial solution pH and common coexistence of nitrate (NO₃⁻) and chloride, on perchlorate removal in the BMIEX process; and (4) verify that PRB can be successfully loaded onto the surface of the MIEX resin to form biofilms by analyzing the microbial community of the biofilm.

2. Materials and methods

2.1. Enrichment culture

Anaerobically digested sludge was obtained from a local municipal wastewater treatment plant (Shanghai, China), which was washed with tap H₂O to reduce the coarse solid particles in the sludge. The concentrated sludge was subsequently collected and served as the seed microorganisms for the enrichment of perchlorate-reducing cultures. The fresh medium had a pH of 7.0 ± 0.2 (adjusted using phosphate buffer) and contained the following components (analytical grade) per liter of tap H₂O: 1.50 g of K₂HPO₄·3H₂O, 0.80 g of NaH₂PO₄·H₂O, 0.50 g of NH₄H₂PO₄, 0.05 g of MgSO₄·7H₂O, and 1 mL of a trace element solution (3.0 g of Na₂EDTA, 4.0 g of FeSO₄·7H₂O, 2.0 g of ZnSO₄·7H₂O, 0.2 g of CuSO₄·5H₂O, 1.0 g of MnCl₂·4H₂O, 1.0 g of CaCl₂·2H₂O, 0.4 g of Na₂MoO₄·2H₂O, 0.4 g of CoCl₂·6H₂O, 0.1 g of NiCl₂·6H₂O, and 0.6 g of H₃BO₃ per liter). The medium was continuously sparged with pure N₂ for 25 min prior to use. Sterile stock solutions of sodium acetate (CH₃COONa, 10 g L⁻¹) and sodium perchlorate (NaClO₄, 10 g of ClO₄⁻ per L) were dissolved in ultrapure water and maintained anaerobically as described above.

Prior to the inoculation of the MIEX resin, the perchlorate-reducing cultures were acclimated and enriched in two 1000 mL autoclaved serum bottles containing 800 mL of medium; the medium was sparged with pure N₂ for 15 min, and the bottles were tightly closed with butyl rubber stoppers. Perchlorate and acetate were then added to the medium at final concentrations of 100 mg L⁻¹ and 2.5 g L⁻¹, serving as the electron acceptor and donor, respectively. The acetate was provided in excess of the mass needed to fully reduce all the perchlorate to chloride. Then, the culture was incubated at 35 °C in the dark in order to avoid any oxygenic photosynthesis and keep the cultures anaerobic. Periodically, approximately 500 mL of the supernatant was decanted and replaced with fresh medium containing the same contents. When both consortiums were able to reduce all of the 100 mg L⁻¹ of ClO₄⁻ in 2 h, they were chosen as the inoculum for subsequent experiments.

2.2. Inoculation of the MIEX resin

The MIEX resin, used as the carrier in this study, was obtained from Orica Watercare of Victoria, Australia. Prior to use, the resin was washed with deionized H₂O; sequentially rinsed with 1 mol L⁻¹ HCl, deionized H₂O and 1 mol L⁻¹ NaOH, and then air-dried at 40 °C. The MIEX resin was then inoculated as described above in two 1000 mL serum bottles, each containing 150 mL of dry MIEX resin and 800 mL of the enriched mixed consortium culture (EMCC). The EMCC consisted of the enrichment cultures and the fresh medium spiked with 50 mg L⁻¹ acetate and 10 mg L⁻¹ perchlorate. A nitrogen purge lowered the DO concentration of the EMCC to approximately 0.30 mg L⁻¹. The content of the bottles was continuously stirred by a thermostat orbital shaker at 100 rpm and 35 °C. During each cycle of batch cultivation (about 2 days), the MIEX resin was quickly settled by gravity, and the old EMCC was then decanted and replaced by the same volume of fresh EMCC. This fill-and-draw culture was continuously operated for 14 feed cycles before the BMIEX resin was formed.

2.3. Perchlorate removal in sequencing-batch reactors

2.3.1. Cycle batch experiments. After 14 feed cycles, the BMIEX resin was removed from the bottles and then slightly flushed with fresh oxygen-free medium several times to remove the excess biomass in the gaps between the BMIEX resin particles. Next, four cycles of batch experiments were conducted to determine the efficacy and evaluate the feasibility of the destruction of perchlorate by BMIEX. For the cycle batch tests, 1 mL of cleaned BMIEX resin, virgin MIEX resin and PRB culture were transferred into 250 mL autoclaved serum bottles, separately. The desired amounts of CH₃COONa and ClO₄⁻ (Table 1) and 200 mL of medium were subsequently added to the bottles. All bottles were sparged with pure N₂ (g) to maintain anaerobic conditions (DO ≤ 0.3 mg L⁻¹) prior to the start of the experiments. All tests were performed in a thermostat orbital shaker at 150 rpm and 35 °C. Moreover, all the cycle batch tests were continuously operated according to the same conditions described in Table 1. When a cycle was complete, all of the resins, including the BMIEX and virgin MIEX resins, were directly used in the next cycle without any regeneration. Every cycle test lasted 12 h, and samples were taken from the bottles at different time intervals. All tests, including the experiments described below, were performed in triplicate unless otherwise stated.

Table 1 Operational conditions for four cycles of batch experiments

Name	Volume, mL	Cycle 1–4
		CH3COONa dosage	Initial pH	T, °C	Initial DO, mg L^−1	Added ClO4^−, mg L^−1
BMIEX resin	1	40 mg L^−1	6.85	35	0.19	10
BMIEX resin	1	0	6.85	35	0.23	10
MIEX resin	1	0	6.85	35	0.17	10
Sludge	1	40 mg L^−1	6.85	35	0.16	10

---

2.3.2. Dissolved oxygen tests. Batch BMIEX tests were conducted to study the perchlorate removal in the presence of oxygen and verify that microbial activity resulted in perchlorate removal. For each DO test (Table 2), 1 mL of BMIEX resin and 200 mL of medium with perchlorate and acetate concentrations of 10 mg L⁻¹ and 40 mg L⁻¹, respectively, were added into the serum bottle, and the pH was maintained at 6.9 ± 0.2. The content of the bottles was sparged with pure N₂ for different durations (8, 2, 1 and 0 min) to obtain various initial DO concentrations (0.24, 2.07, 5.22 and 8.47 mg L⁻¹, respectively). Batch tests were performed in a thermostat orbital shaker at 150 rpm and 35 °C and lasted for at least 18 h. The samples were taken from the bottles at specific time intervals, and the residual DO in the solution was monitored simultaneously. The detailed experimental procedures are described in Section 2.3.1 and Table 2.

Table 2 Experimental procedures for the sequencing batch reactor^a

Factors	Added ClO4^−, mg L^−1	CH3COONa dosage	DO, mg L^−1	Initial pH	Added NO3^−, mg L^−1	Added Cl^−, mg L^−1
a Horizontal lines represent no added.
DO	10	40	0.24–8.47	6.8	—	—
pH	10	40	0.24	4.5–10	—	—
NO3^−	10	40	0.24	6.8	0–40	—
Cl^−	10	40	0.24	6.8	—	0–1000

---

2.3.3. Effect of media pH. To investigate the effect of pH on perchlorate degradation by the BMIEX resin, media containing 10 mg L⁻¹ perchlorate and 40 mg L⁻¹ acetate were adjusted to different pH levels ranging from 4.5 to 6.8 and 8.0 and 10, respectively. The initial pH was adjusted by adding the required quantity of 1.0 M HCl or 1.0 M NaOH. The bottles were incubated in a shaking incubator at 150 rpm and 35 °C. The detailed experimental procedures are similar to the method described in Section 2.3.2 and Table 2.

2.3.4. Competing ion tests. The capability of the BMIEX resin to removal perchlorate in the presence of other competing anions (that act as either potential competitive electron acceptors and preferential adsorbates against perchlorate or as exchangeable ions and reaction products) such as nitrate and chloride was also analyzed using two pairs of commonly coexisting ion systems (ClO₄⁻–NO₃⁻, ClO₄⁻–Cl⁻). In the ClO₄⁻–NO₃⁻ system, nitrate was added to result in a concentration ranging between 0 and 40 mg L⁻¹. Similarly, effect of chloride on the removal of perchlorate by the BMIEX resin was investigated at five different values ranging from 0 to 1000 mg L⁻¹ at pH 6.9 ± 0.2 of the media. As such, the detailed experimental procedures are similar to the method described in Section 2.3.2 and Table 2.

2.4. Chemical analyses

Samples from the reactors were obtained using syringes, and all samples were centrifuged to separate suspended solids. Then, all samples were filtered through 0.22 μm hydrophilic polyethersulfone (PES) syringe reactors (Anpel, Shanghai, China) and stored at 4 °C. Perchlorate, nitrate and chloride were analyzed using a Dionex (Sunnyvale, CA) ICS-2000 ion chromatograph equipped with a Dionex IonPac AS20 analytical column, an AG20 guard column, a 150 μL sample loop and a self-regenerating suppressor. Perchlorate was measured with 35 mmol L⁻¹ KOH as the eluent at a rate of 1 mL min⁻¹, and the detection limit was approximately 4 μg L⁻¹. Dissolved organic carbon (DOC) was measured on a total organic carbon analyzer (TOC-L, CPH CN200, Shimadzu) with a detection limit of 50 μg L⁻¹. Moreover, DO was measured on a galvanic oxygen sensor (WTW CellOx 325 with Oxi 340i, Weilheim, Germany) with a probe that had been fitted with a rubber gasket, and the detection limit was 0.1 mg L⁻¹. The solution pH was measured with a Mettler Toledo pH meter.

2.5. Analysis of the biofilm community

The biofilm biomass attached to the carriers was separated by ultrasonic treatment and centrifuged to isolate the DNA at the end of the experiment. Genomic DNA was extracted from the biofilm sample using the E.Z.N.A. DNA Isolation Kit (OMEGA Biotec, USA) according to the manufacturer's instructions and stored in a refrigerator at −20 °C. The diversity of the biofilm community was analyzed by high-throughput sequencing on a massively parallel 454 GS-FLX Titanium sequencer (Roche 454 Life Sciences, Branford, CT, USA). Taxonomic classification of the sequences was performed using the RDP Classifier (version 2.2) with a set confidence threshold of 70%. The community diversity and relative bacterial community richness were obtained based on OTUs (OTU diversity) and on reads (OTU abundance), respectively, in Venn diagrams.

3. Results & discussion

3.1. Perchlorate removal during the 4-cycle batch tests using the BMIEX resin, virgin MIEX resin and PRB culture

A series of batch cycle tests using the BMIEX resin, virgin MIEX resin and PRB culture were performed to evaluate the feasibility of the BMIEX method to remove perchlorate. A batch test method was employed here because it is easy to operate and quick to respond. Summaries of the conditions for these experiments are given in Table 1. For the 4-cycle batch tests, the time course of residual perchlorate in the bulk solution is shown in Fig. 1. In cycle 1, shown in Fig. 1(a), the perchlorate removal by the virgin MIEX resin rapidly reached equilibrium in 1 h, with an overall removal efficiency of 89.6%. Perchlorate removal using the BMIEX resin with/without 40 mg L⁻¹ acetate added as a carbon source (expressed as BMIEX with/without C system) had a similar pattern and rate (average ClO₄⁻ removal rate of 9.32 mg L⁻¹ h⁻¹) during the first 1 h. However, a subsequent continuous slight decrease instead of an equilibrium in the perchlorate concentration from 263 to 0 μg L⁻¹ was observed between 1 and 12 h in the BMIEX with C system, and the BMIEX without C system exhibited similar trends over 11 h, with a 95.23% removal efficiency. The initial rapid decrease in the perchlorate concentration is mainly due to adsorption kinetics, and the subsequent slight decrease mainly results from microbial degradation in the BMIEX with C system. In cycle 2, shown in Fig. 1(b), the perchlorate removal rate decreased significantly, and the overall removal efficiencies in the MIEX and BMIEX without C systems were 78% and 84%, respectively. However, in the BMIEX with C system, the perchlorate removal rate was rapid during the first 1 h and slowed between 1 and 12 h, with undetected perchlorate. In cycles 3 and 4, further decreases in the perchlorate removal rates in the virgin MIEX and BMIEX without C systems were observed, which are signs of MIEX and BMIEX resin failure (Fig. 1(c) and (d)). In contrast, the high removal rate of perchlorate (99.0% and 95.6% in cycles 3 and 4, respectively) in the BMIEX with C system demonstrates the durability of the BMIEX resin combined with added carbon to remove perchlorate from water. Moreover, Fig. 1 shows that the PRB culture system (the biomass in the PRB culture is much higher than that in the BMIEX with C system) had a lower but stable removal rate of perchlorate (average removal rate of 46.7%). All these results suggest that the removal of perchlorate from water using the BMIEX resin was effective and the efficiency of the BMIEX resin is higher than that of either the MIEX resin or PRB culture alone. Here, the MIEX resin provides an attachment surface for the PRB and protects them from shock loadings of toxic or inhibitory materials; the microorganisms bioregenerate the MIEX resin in return. The BMIEX treatment can be divided into three phases similar to the BAC treatment processes. The first phase involves that perchlorate was rapidly concentrated from the water to the BMIEX resin, when the rate of adsorption considerably surpasses the biodegradation rate. The second phase is desorption–biodegradation equilibrium, when the rates of both desorption and biodegradation are comparable. The third corresponds to the primary biodegradation in which the biodegradation rate surpasses the desorption rate, resulting in regeneration of the resin.

Fig. 1 Time courses of the residual perchlorate (ClO₄⁻) concentration in the aqueous phase during the 4-cycle batch tests: cycle 1 (a), cycle 2 (b), cycle 3 (c), and cycle 4 (d) at an initial perchlorate concentration of 10 mg L⁻¹.

A comparison of the DOC consumption before and after the 4-cycle batch tests is shown in Fig. 2. Almost complete DOC removal (∼12 mg L⁻¹ DOC out of 16.3 mg L⁻¹ acetate as added C) was observed in the effluent of every BMIEX with C system, along with a high removal of perchlorate. In comparison, DOC was also consumed in the PRB culture system due to microbial degradation as follows: CH₃COOH + ClO₄⁻ → 2CO₂ + Cl⁻ + 2H₂O.²⁶ In the BMIEX with C system, PRB attached on the BMIEX resin reduced perchlorate along with acetate as the carbon source, leading to the significant decrease in DOC. Additionally, a comparison of Fig. 1 and 2 indicates that the perchlorate sorption capacity of the MIEX resin was quickly exceeded without any regeneration or PRB degradation. The removal effect of perchlorate using the BMIEX resin with C system was much better than that of the BMIEX resin without C system, which further demonstrates that microbial perchlorate reduction can prolong the service life of the BMIEX resin and that an additional carbon source is needed for microbial metabolism. Besides, the removal of perchlorate in the BMIEX resin without C system was slightly better than that in the MIEX resin system, which demonstrates that the BMIEX resin could adsorb and accumulate background organic matter from the water for the biofilm used as an electron donor. Overall, the BMIEX method effectively removed perchlorate from the water, and the BMIEX resin achieved perchlorate destruction rather than a simple enrichment.

Fig. 2 Comparison of DOC consumption before and after the batch tests: cycle 1 (a), cycle 2 (b), cycle 3 (c), and cycle 4 (d). The initial acetate dosage was 40 mg L⁻¹ in the BMIEX with C and PRB culture systems, respectively.

3.2. Effect of DO on perchlorate removal

PRB preferentially utilize O₂ over perchlorate when both electron acceptors are provided simultaneously.²⁷ Therefore, the effects of the initial DO concentration on perchlorate removal in the BMIEX with C system were investigated over a wide range (0.24–8.27 mg L⁻¹) here. As shown in Fig. 3, perchlorate was almost completely removed in 18 h at all tested DO levels. Comparatively, perchlorate removal was the most rapid with an initial DO concentration of 0.24 mg L⁻¹, and the removal efficiency reached 96.8% within 2 h. Only 60.7% of perchlorate was removed in the same time frame with an initial DO concentration of 8.47 mg L⁻¹, along with a decrease in DO from 8.27 to 0.98 mg L⁻¹, but an increase in perchlorate removal was subsequently observed again as the DO was consumed. Perchlorate removal by the BMIEX resin was significantly slowed by DO, that is, oxygen was preferred as an electron acceptor over perchlorate by the PRB, leading to an adverse effect on perchlorate removal. Furthermore, when the initial DO was decreased to 5.22 and 2.07 mg L⁻¹, the corresponding removal of perchlorate within 2 h was respectively enhanced to 86.5% and 92.4%, indicating the adverse effect of DO in removal of perchlorate. At the end of the tests, the removal percentages of perchlorate in 0.24, 2.07, 5.22 and 8.27 mg L⁻¹ DO were 100, 98.9, 98.7 and 98.6%, respectively. These results can be interpreted as evidence that the relationship between the MIEX resin and PRB biofilm in this BMIEX system is complementary rather than mutually independent and that PRB play an important role in the destruction of perchlorate by BMIEX. Similar results regarding the reaction of BAC were obtained by Chudyk,²⁸ who reported that oxygen is preferred by PRB as an electron acceptor over perchlorate and that high dissolved oxygen concentrations completely inhibit perchlorate reduction.

Fig. 3 Time courses of the residual perchlorate (ClO₄⁻) concentration and dissolved oxygen (DO) in the aqueous phase at 4 initial DO concentrations: 0.24 mg L⁻¹ (a), 2.07 mg L⁻¹ (b), 5.22 mg L⁻¹ (c), 8.47 mg L⁻¹ (d). All treatments received 40 mg L⁻¹ acetate and 10 mg L⁻¹ perchlorate, respectively.

3.3. Effect of medium pH

pH is an important factor affecting microbial reactivity. A high removal efficiency of perchlorate can be obtained using virgin MIEX resin at a pH of 4.0–9.0,²⁵ and most PRB require a neutral pH of approximately 6.8–7.2 for growth and optimal perchlorate reduction.²⁷ The effect of the medium's pH on the extent of perchlorate removal by the BMIEX resin was evaluated at pH values ranging from 4.5 to 10. Fig. 4 shows that the maximum removal of perchlorate was observed at pH 6.8, for which perchlorate decreased substantially from an initial concentration of 10 mg L⁻¹ to 57 μg L⁻¹ (99.4% removal) within 4 h; complete removal of perchlorate was observed within 12 h. However, the perchlorate reduction activity of the BMIEX process decreased sharply at pH 4.5 and 10 with respective removal efficiencies of 86.8% and 84.7%, which indicates that the removal of perchlorate was mainly due to adsorption by the BMIEX resin rather than reduction by PRB. Moreover, no substantial difference was noted for pH 8 compared with pH 10. The effective pH for the BMIEX process was at a neutral range, and promising ClO₄⁻ removal was achieved in cultures of pH 6.8.

Fig. 4 Time courses of the residual perchlorate (ClO₄⁻) concentration in the aqueous phase at different pH values. All treatments received 40 mg L⁻¹ acetate.

3.4. Effects of commonly coexisting ions

Nitrate is commonly present in most perchlorate-contaminated water systems at concentrations several orders of magnitude greater than that of perchlorate.²⁷ Additionally, nitrate, as an outstanding competitor of ClO₄⁻, was observed to cause obvious lags in biological perchlorate reduction^29,30 and had certain inhibiting effect on perchlorate adsorption by MIEX resin.²⁵ The effect of nitrate on perchlorate removal was examined to evaluate the feasibility of applying the BMIEX process to water containing different concentrations of nitrate, which is shown in Fig. 5(a). Perchlorate removal rate decreased from 96.8% to 92.5% during the first 1 h when nitrate concentration varied from 0 to 40 mg L⁻¹. As mentioned above, the initial rapid decrease in the perchlorate concentration within the first hour was mainly attributed to the chemical adsorption kinetics of the BMIEX resin. The results demonstrated that increasing nitrate concentrations had not obvious effect on the adsorption of perchlorate by the BMIEX resin. Notably, the presence of 40 mg L⁻¹ NO₃⁻ completely inhibited further reduction of ClO₄⁻ within 1–18 h, and ClO₄⁻ reduction in the presence of 20 mg L⁻¹ and 5 mg L⁻¹ NO₃⁻ did not commence until all NO₃⁻ in the microcosms had been depleted at 10 and 6 h, respectively. The observations above indicated that the increase of nitrate concentration would lead to a longer lag even complete inhibition. Complete removal of perchlorate was observed during the initial 12 h without a lag in the absence of nitrate. These results demonstrated that the BMIEX resin tended to reduce nitrate in preference to perchlorate when both electron acceptors were provided simultaneously. The negative effect of nitrate on the perchlorate removal by BMIEX mainly resulted from biofilm reduction, as compared to adsorption.

Fig. 5 Time courses of the residual perchlorate (ClO₄⁻) concentration in the aqueous phase in the presence of nitrate (a) and chloride (b) at different concentrations. All treatments received 40 mg L⁻¹ acetate.

Chloride, as a common ion, is an exchangeable product ion of the MIEX resin and a reaction product of biological perchlorate reduction. As Fig. 5(b) shows, the perchlorate removal efficiency of the BMIEX process gradually decreased with a conspicuous increase in the initial Cl⁻ concentration, indicating that increasing chloride concentrations can slightly inhibit perchlorate removal by chemical adsorption and microbial reduction in the BMIEX process.

3.5. Microbial community structure analysis of the biofilm

The microbial community composition was analyzed based on clone library results (Table 3). The phylum Proteobacteria (alpha, beta, gamma and delta) was clearly predominant, representing approximately 60.23% of the total bacteria in the biomass sample. Moreover, among all the Proteobacteria, the abundances of α-, β-, γ- and δ-Proteobacteria were approximately 21.04%, 12.44%, 24.33% and 0.71%, respectively. According to prior research, most identified dissimilatory PRB phylogenetically belong to the α, β and γ subclasses of the Proteobacteria.³¹ These results demonstrate that attachment to the MIEX resin occurred.

Table 3 Microbial community compositions by relative abundance (%) in the biomass sample collected from the BMIEX resin biofilms

Phylum-level taxon	Class-level taxon		Relative abundance (%)
Proteobacteria			60.23
		Alphaproteobacteria	21.04
		Betaproteobacteria	12.44
		Gammaproteobacteria	24.33
		Deltaproteobacteria	0.71
		Proteobacteria_unclassified	1.71
Actinobacteria			11.55
Chloroflexi			8.02
Acidobacteria			7.40
Bacteria_unclassified			3.44
Firmicutes			2.29
Cyanobacteria			3.44
Other			3.63

---

4. Conclusions

A method was developed that combined biological and MIEX resin-adsorption treatments, expressed as BMIEX. The BMIEX method was effective at removing perchlorate from water, and an additional carbon source was needed for microbial metabolism. Perchlorate removal by the BMIEX resin was significantly slowed by DO. A neutral pH range was optimal, and maximum perchlorate removal was obtained at pH 6.8. The coexistence of NO₃⁻ and Cl⁻ can inhibit the rate and extent of perchlorate removal, especially NO₃⁻. The dominant bacteria in the consortium were Proteobacteria (60.23%), with most identified as dissimilatory PRB.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 51178321; No. 51208364), the National Major Project of Science & Technology Ministry of China (No. 2012ZX07403-001; No. 2008ZX07421-002), the research and development Project of Ministry of Housing and Urban-Rural Development (No. 2009-K7-4) and the Fundamental Research Funds for the Central Universities (0400219279).

References

1. R. Nerenberg, Y. Kawagoshi and B. E. Rittmann, Water Res., 2006, 40, 3290–3296.
2. B. Gu, G. M. Brown and C.-C. Chiang, Environ. Sci. Technol., 2007, 41, 6277–6282.
3. Y. C. Choi, X. Li, L. Raskin and E. Morgenroth, Water Res., 2008, 42, 3425–3434.
4. S. H. Lee, J. H. Hwang, A. N. Kabra, R. A. I. Abou-Shanab, M. B. Kurade, B. Min and B. H. Jeon, Environ. Sci. Pollut. Res., 2015, 22, 18839–18848.
5. J. J. Li, M. M. Gao, G. Zhang, X. H. Wang, S. G. Wang, C. Song and Y. Y. Xu, Bioresour. Technol., 2015, 177, 74–79.
6. R. Nerenberg, B. E. Rittmann and I. Najm, J. - Am. Water Works Assoc., 2002, 94, 103.
7. S. K. Chaudhuri, S. M. O'Connor, R. L. Gustavson, L. A. Achenbach and J. D. Coates, Appl. Environ. Microbiol., 2002, 68, 4425–4430.
8. D. Ucar, E. U. Cokgor and E. Sahinkaya, Environ. Technol., 2016, 37, 183–191.
9. H. W. Ryu, S. J. Nor, K. E. Moon, K. S. Cho, D. K. Cha and K. I. Rhee, Bioresour. Technol., 2012, 103, 279–285.
10. U.S. EPA, Fed. Regist., 2011, 76(29), 7762–7767.
11. R. Nerenberg, Y. Kawagoshi and B. E. Rittmann, Water Res., 2008, 42, 1151–1159.
12. J. D. Coates and L. A. Achenbach, Nat. Rev. Microbiol., 2004, 2, 569–580.
13. C. Wang, L. Lippincott, I. H. Yoon and X. Meng, Water Res., 2009, 43, 127–136.
14. A. K. Venkatesan, M. Sharbatmaleki and J. R. Batista, J. Hazard. Mater., 2010, 177, 730–737.
15. H. Choi and J. Silverstein, J. Hazard. Mater., 2008, 159, 440–445.
16. S. Ebrahimi and D. J. Roberts, Sep. Purif. Technol., 2015, 146, 268–275.
17. X. Meng, D. A. Vaccari, J. Zhang, A. Fiume and X. Meng, Environ. Sci. Technol., 2014, 48, 1541–1548.
18. J. Ren, W. Yang, M. Hua, B. Pan and W. Zhang, Bioresour. Technol., 2013, 142, 701–705.
19. Z. Ren, X. Xu, B. Gao, Q. Yue and W. Song, J. Colloid Interface Sci., 2015, 459, 127–135.
20. M. Sharbatmaleki, R. F. Unz and J. R. Batista, Water Res., 2015, 75, 1–10.
21. A. K. Venkatesan and J. R. Batista, Biorem. J., 2011, 15, 1–11.
22. C. Wang, L. Lippincott and X. Meng, Water Res., 2008, 42, 4619–4628.
23. M. Sharbatmaleki and J. R. Batista, Water Res., 2012, 46, 21–32.
24. Y. Zhu, N. Gao, Q. Wang and X. Wei, Colloids Surf., A, 2015, 468, 114–121.
25. Y. Tang, S. Liang, H. Guo, H. You, N. Gao and S. Yu, Colloids Surf., A, 2013, 417, 26–31.
26. S. J. Nor, S. H. Lee, K.-S. Cho, D. K. Cha, K. I. Lee and H. W. Ryu, Bioresour. Technol., 2011, 102, 835–841.
27. N. Bardiya and J. H. Bae, Microbiol. Res., 2011, 166, 237–254.
28. Y. Song and B. E. Logan, Water Res., 2004, 38, 1626–1632.
29. D. Xie, H. Yu, C. Li, Y. Ren, C. Wei and C. Feng, Electrochim. Acta, 2014, 133, 217–223.
30. Y. Tang, H. Zhao, A. K. Marcus, R. Krajmalnik-Brown and B. E. Rittmann, Environ. Sci. Technol., 2012, 46, 1598–1607.
31. U. Michaelidou, L. A. Achenbach and J. D. Coates, in Perchlorate in the Environment, Springer, 2000, pp. 271–283.

---