Batch laboratorial experiments were conducted to evaluate the potential of Fe0, Cu0 and sulfate-reducing bacteria (SRB) for reduction and removal of sulfate and heavy metals from synthetic acid mine drainage. The variation in solution pH indicates that the Fe0/Cu0 bimetallic system provided favorable conditions for SRB growth and sulfate reduction. When the SO42− concentration of wastewater was 3,000 mg/L, the SO42− removal efficiency was 51.6% for the SRB system, 76.3% for the Fe + SRB system, and 92.0% for the Fe/Cu + SRB system. Moreover, Pb2+, Cu2+ and Zn2+ ions were completely removed. The results demonstrate that the Fe/Cu + SRB system had apparent advantages over the SRB system, especially at low pH. This study demonstrates that an Fe/Cu + SRB system could be a promising technology for treating wastewater containing high concentrations of sulfate and heavy metals.
Acid mine drainage (AMD), generated by the biological oxidation of metal sulfides to metal sulfates in mine wastes, is characterized by low pH and high concentrations of dissolved metals and sulfates (Peng et al. 2017). AMD is extremely toxic to aquatic organisms because of its strong acidity and the solubilization of metals (Peng et al. 2017). Moreover, the sulfates can be reduced to sulfides by sulfate-reducing bacteria (SRB) under anaerobic conditions, leading to air pollution and corrosion of metal pipes. Thus, AMD represents a challenging environmental issue in the mining industry worldwide.
There exist various methods for treating sulfates in wastewater, including activated carbon adsorption (Salman 2018), biological treatment (Singh et al. 2011; Rodriguez et al. 2016), reverse osmosis (Al-Bastaki 2004), etc. The physical processes have disadvantages such as high infrastructure costs and high operating costs. Therefore, biological processes have been employed as an alternative to treat sulfate-containing wastewater (Singh et al. 2011; Rodriguez et al. 2016). Sulfates are non-toxic to organisms, but their reduction products (namely sulfides) have a highly poisonous effect on microbes including SRB (Singh et al. 2011). Thus, treating AMD by biological processes alone may be hampered by the presence of high concentrations of sulfides.
Zero-valent iron (ZVI, Fe0), a strong reducing reagent (E0 = −0.447 V), in aqueous solution, acts as an electron donor under certain conditions and can carry out oxidation-reduction reactions without any external energy input. ZVI is expected to lower the oxidation-reduction potential (ORP), which is beneficial for the formation of an enhanced anaerobic environment for SRB (Kumar et al. 2015). Additionally, Fe0 can release from Fe0 to Fe2+ (Fe0 + 2H+ = Fe2+ + H2) and thus buffer the solution pH (Liu et al. 2015) and reduce the toxicity of H2S by the formation of FeS precipitation (Zhang et al. 2011). Fe0 can also lower mobility and toxicity of some heavy metals (Cu2+ and Cr6+) through reduction behavior. Therefore, Fe0 packed anaerobic bioreactors have been developed to improve the treatment performance of sulfate-containing wastewater, resulting in better sulfate reduction and metal removal (Singh et al. 2011; Zhang et al. 2011; Ayala-Parra et al. 2016). The application of Fe0 in anaerobic bioreactors could maintain higher organic removal and higher microbe density (Zhang et al. 2011). Moreover, it was found that Fe0 addition enhanced sulfate reduction and generated alkalinity (Ayala-Parra et al. 2016).
However, the reactivity of Fe0 tends to decline with time because of the accumulation of hydroxide precipitates and other corrosion products on the metal surface. It has been demonstrated that Cu0 addition could provide high reaction potential and improve reduction capacity of Fe0, resulting in a higher pollutant removal efficiency even in a neutral pH range (Lai et al. 2014). Thus, this study proposed an Fe/Cu bimetallic packed anaerobic system to improve sulfate removal by SRB from AMD. The validity and applicability of the system is tested and compared under different conditions.
MATERIALS AND METHODS
Iron shavings (Fe0) and copper shavings (Cu0) were collected from a local machining factory. The shavings used had a purity of greater than 99% and a grain size range of 14–18 mesh. The shavings (1 g) were mixed with 10 mL of 10% (v/v) HCl solution for 2 min and were then rinsed with nitrogen-sparged deionized water, and then rinsed in acetone. The shavings were dried with N2 inside the anaerobic glove box before use.
The mixed SRB population was enriched from activated sludge obtained from the anaerobic digester at a local sewage treatment plant. For this, 5 g anhydrous Na2SO4 was added into a 1-L bottle containing 1 L of anaerobic sludge. After 7 days of incubation at 37 °C, 50 mL of sludge was transferred to another 1-L bottle containing 950 mL of autoclaved Postgate's B medium (Postgate 1984), and then incubated at 37 °C for another 7 days. This procedure was repeated 7 times to enrich SRB. The resultant culture was centrifuged (8,000 g, 10 min), washed twice with physiological saline, and centrifuged again as indicated above to collect the pellet (namely the final consortium). Light microscopy and scanning electron microscopy (SEM) showed that motile vibrios dominated the population. The presence of SRB was further confirmed by the observation of the precipitation of black iron sulfide. The SRB consortium was stored under N2 at 4 °C, at the College of Life Science, Longyan University. The 16S rRNA gene based clone libraries demonstrated that the SRB consortium mainly contained Desulfomicrobium spp., Desulfobulbus spp., and Desulfovibrio spp.
In this work, an artificial AMD was prepared by using Na2SO4, Pb(NO3)2, Cu(NO3)2, and ZnCl2. The stated concentrations of SO42−, Pb2+, Cu2+ and Zn2+ ions were 3,000, 50, 50 and 50 mg/L, respectively, in the wastewater prepared (Peng et al. 2017). To provide inorganic nutrient and organic substrate for SRB, the AMD was supplemented with K2HPO4 (0.5 g/L), NH4Cl (1.0 g/L), CaCl2 (0.1 g/L). Chemical oxygen demand (COD) was adjusted as required with glucose.
Batch tests were conducted in 1-L glass bottles, to test the effect of various parameters on the removal of sulfates. The bottles had one inlet for probe insertion and two outlets for sampling and gas discharge. The bottles were connected with 500 mL aspirated bottles (gas holder) filled with 2.0 M NaOH solution to trap the H2S generated during incubation.
SRB consortium (about 5 g) was transferred to bottles with 900 mL artificial AMD. The headspace was flushed with N2/CO2 gas (80:20, v/v) to exclude oxygen and bottles were sealed with butyl rubber septa to exclude oxygen permeation. Fe0 and Cu0 were added to the mixture at a final concentration of 2 and 0.2 g/L, respectively. The bottles were then incubated at 100 rpm and 37 °C. There were four treatments with three replications as follows: (1) Fe/Cu; (2) SRB; (3) Fe + SRB; (4) Fe/Cu + SRB. In each treatment, the volume of AMD was identical (900 mL).
Liquid samples were obtained from each bottle using a plastic syringe at a pre-determined time. The supernatant was collected for analysis after centrifugation at 8,000 rpm for 10 min. Sulfide, sulfite, and sulfate were determined by ion chromatography (C196-E039A, Shimadzu, Japan) with a Shim-pack IC-A3 chromatographic column. Metal ions were measured using atomic absorption spectrophotometry (Shimadzu AA6300G, Kyoto, Japan) with a graphite furnace atomizer (GFA-EX7i). Microbial populations were enumerated following the completion of the batch experiments. Relative populations of SRB and iron reducing bacteria (IRB) were determined using the most probable number (MPN) technique (Cochran 1950).
RESULTS AND DISCUSSION
Effect of Fe0 and Cu0 on sulfate reduction
The performance of the systems was analyzed by taking into account the efficiency of sulfate reduction when the COD concentration of AMD was 3,600 mg/L, and the initial pH was 7.0. The effect of Fe0 and Cu0 on sulfate reduction is shown in Figure 1. As shown, no sulfate reduction occurred in the case of Fe/Cu. This indicates that sulfate could not be reduced by Fe0, and the adsorption of sulfate by Fe0 or Cu0 was negligible. The standard Gibbs free energy change (ΔG0) for the reaction 4Fe0 + SO42− + 8H+ = 4Fe2+ + S2− + 4H2O is about 1,430 KJ/mol > 0. Thus, this reaction cannot take place spontaneously.
H2 can then serve as an electron donor for stimulating the anaerobic biotransformation of reducible contaminants. Additionally, there is evidence for direct electron transfer from Fe0 to microbes based on SRB accelerating the rate of Fe0 corrosion beyond the abiotic rate (Venzlaff et al. 2013). Moreover, adding Fe0 could reduce the toxicity of sulfides to microbes by the formation of FeS precipitation (Zhang et al. 2011).
Figure 1 also demonstrates that adding Cu0 can further promote sulfate reduction by SRB relative to Fe + SRB. Cu0 can improve the reduction capacity of Fe0 due to the high reaction potential, and higher pollutant removal efficiency could be obtained even in the neutral pH range (Fan & Ma 2009). Two types of catalytic mechanisms of the bimetallic reductants have been proposed (Lai et al. 2014): (a) indirect reduction by the atomic hydrogen absorbed on the surface of bimetallic reductants, and the transition metal additives can facilitate the generation of surface-bond atomic hydrogen, (b) direct reduction on the catalytic activity site by accepting electrons from the oxidation of Fe0, and the transition metal additives can increase the oxidation of Fe0 through the formation of infinite galvanic cells.
Effect of COD/SO42− ratio on sulfate reduction
The COD mass used for the desulfurization was calculated based on stoichiometry that 0.67 g COD is required to reduce 1 g of SO42−; meanwhile, 1.042 g of alkalinity is produced (as CaCO3) (Vossoughi et al. 2003). Thus, in subsequent experiments, the effect of COD/SO42− ratio on SO42− reduction was investigated by varying the COD concentration of wastewater, while the SO42− concentration was kept at 3,000 mg/L.
Figure 2 shows the plots of the COD/SO42− ratio and sulfate removal in the bottles after 60 days of incubation. From data presented in Figure 2 it can be observed that the COD/SO42− ratio has an apparent effect on sulfate reduction. The removal efficiency of SO42− significantly increased with increasing the COD/SO42− ratio from 0.3 to 1.2. This demonstrates that the wastewater COD could not satisfy the SRB requirement for growth and sulfate reduction at low COD/SO42− ratio, and SO42− could not be efficiently reduced to sulfide by SRB because of the deficiency of electron donors. When the COD/SO42− ratio was varied between 1.2 and 2.0, sulfate removal remained relatively stable at 51.6–92.2%. In this study, a similar situation was observed among various systems: namely higher sulfate reduction in Fe0-amended systems did not result in a higher organics requirement. This indicates that cathodic H2 could serve as an electron donor for sulfate reduction by SRB (Equation (1)).
Several studies have been conducted to find the optimum COD/SO42− ratios for SRB to reduce sulfate but the results were diverse. The best COD/SO42− ratio was found to be 5.0 with sludge as the organic substrate (Al-Ani 1994), whereas SRB was predominant for a ratio below 1.7 using natural or synthetic substrates (Prasad et al. 1999). When complex organics were used as substrate, higher optimum ratios were likely obtained because not all the carbon present was utilized by SRB. Glucose is an excellent carbon source for microbes, thus a low COD/SO42− ratio was demonstrated in this study.
Effect of initial pH on sulfate reduction
The effect of initial pH on sulfate reduction was tested with pH decreasing from 7.0 to 3.0, when the concentrations of SO42− and COD were respectively 3,000 and 3,600 mg/L, and the temperature of the system was 37 °C. Figure 3 shows the changes of sulfate removal and effluent pH in the bottles after 60 days of incubation across the three systems. As shown, initial pH exerted an apparent effect on sulfate removal across various systems. In the SRB system, sulfate removal gradually declined with the initial pH decreasing from 7.0 to 4.0, and an apparent decrease of sulfate removal was observed with the decrease of pH from 4.0 to 3.0 (Figure 3(a)). In contrast, the influence of initial pH on sulfate removal in the Fe + SRB and Fe/Cu + SRB systems was slighter compared with that in the SRB system (Figure 3). The sulfate removal of the SRB, Fe + SRB and Fe/Cu + SRB systems were 32.5%, 63.5% and 78.4% respectively with the initial pH at 4.0. Sulfate removal was only 23.6% in the SRB system with pH 3.0 (Figure 3(a)). Sulfate removal reached 58.2% and 74.3% in the Fe + SRB and Fe/Cu + SRB systems respectively under the same pH condition (Figure 3(b) and 3(c)). The optimal pH for SRB growth was around 6–7, whereas the performance of the Fe0-amended systems was good for sulfate removal when the initial pH was above 4.0.
A similar behavior of the effluent pH influenced by the changed initial pH was observed in the three systems (Figure 3). The effluent pH in the SRB system was above 6.2 with the initial pH above 4.0. When the initial pH decreased to 3.0, the effluent pH was about 5.5 (Figure 3(a)). In the Fe + SRB and Fe/Cu + SRB systems, effluent pH was above 6.5 with the initial pH from 3.0 to 7.0. The result of initial pH effect on sulfate removal and effluent pH suggests that Fe0 could increase the pH of the systems to improve the SRB activity. In this study, the increase of effluent pH after incubation was primarily due to the release of bicarbonate alkalinity during sulfate reduction (Equation (3)), which is beneficial for acid neutralization and pH buffering (Singh et al. 2011). Additionally, in Fe0-amended systems, alkalinity can be generated by anoxic corrosion of Fe0 (Equation (1)) and sulfate reduction (Equation (2)), as both processes involve the consumption of hydrogen ions. Therefore, the pH values of the effluent from the Fe + SRB and Fe/Cu + SRB systems were higher than that from the SRB system.
To confirm the positive effect of Fe0, heavy metals removal was monitored during the first 96 h of reaction in the Fe, SRB, Fe + SRB and Fe/Cu + SRB systems with Pb2+, Cu2+ and Zn2+ concentrations of 50, 50 and 50 mg/L, respectively (Table 1). In the Fe system, the concentration of Zn2+ was almost unchanged, since Zn2+ could not be reduced by Fe0; however, the Pb2+ and Cu2+ removal was almost 100% after 72 h, indicating the reduction and precipitation of Pb2+ and Cu2+. In the SRB, Fe + SRB and Fe/Cu + SRB systems, the removal ratios of Pb2+, Cu2+ and Zn2+ were 100% after 48 h. At the end of experiments, the residual ratios of Fe0 were 34.3% and 28.5% in the Fe + SRB and Fe/Cu + SRB systems, while the residual ratio of Cu0 was 92.6% (data not shown).
|Time (h)||Fe system||SRB system||Fe + SRB system||Fe/Cu + SRB system|
|Time (h)||Fe system||SRB system||Fe + SRB system||Fe/Cu + SRB system|
It was remarkable that the results of SRB, Fe + SRB and Fe/Cu + SRB systems were identical, indicating that the content of sulfides was sufficient for precipitating with metals. The addition of Fe0 could enhance the pH value of the solution and the sulfate reduction capacity of SRB in anaerobic systems, leading to the accelerated removal of heavy metals. Metal-sulfide mineral precipitation and reduction/precipitation were the two dominant mechanisms of metal removal in this study. However, other mechanisms, such as (oxy)hydroxide precipitation, co-precipitation and surface complexation, may be important (Yue et al. 2015). The contribution of individual routes to metal removal will be further explored in the following work.
Figure 4 shows the populations of SRB and IRB in samples collected after 60 days. The number of SRB populations ranged from 5.2 × 106 to 1.4 × 107 MPN/mL, while no evidence of both SRB and IRB growth was observed in the sample collected from the Fe system. Additionally, no IRB was detected in the SRB system, since no Fe0 was applied in this case. Populations of IRB ranged from 3.7 × 104 to 2.3 × 105 MPN/mL, and the MPN value was approximately 2 orders of magnitude lower than that of SRB (Figure 4).
The SRB populations are consistent with calculated sulfate reduction efficiency (Figure 1). Although H2(g) concentrations were not measured, increments in pH were indicative of anaerobic Fe0 corrosion which produced H2(g) (Equation (1)). SRB growth could be enhanced with the addition of H2 as an electron donor (Ayala-Parra et al. 2016). Improved sulfate reduction efficiency in the presence of Fe0 may therefore correspond to higher relative populations of H2 oxidizing SRB.
Sulfur element balance
Sulfate conversion was analyzed with the contents of different forms of sulfur. Figure 5 shows the S balance in the three systems. The proportion of sulfate converted into gas H2S was 20.6–45.4%, whereas aqueous sulfide in the effluent accounted for 6.3–21.5% of influent sulfate. About 5.7–36.8% of sulfate was reduced to S2− and then precipitated with metals such as Fe, Pb, Cu and Zn. About 8–48.4% of sulfate was left in the effluent. Moreover, 3.5–3.8% of sulfate dissipated through other processes such as microbial assimilation, oxidation of sulfide into elemental sulfur, etc. Apparently, the addition of Fe0 greatly reduced the content of aqueous sulfide, which may lessen the inhibitory effect of sulfide on SRB.
A rough economic analysis was performed on the operating costs, such as the costs of reagents and the costs of energy. It should be noted that this analysis is just an approximate tool to differentiate the trends in the operating costs associated with the use of combined treatment. A rigorous economic analysis should consider initial investment, prices at plant scale, maintenance and labor costs and so forth. Table 2 shows that the total reagent and power costs were $0.95/m3. The cost of an advanced and complete treatment is always up to $5–7/m3 for industrial wastewater (Oller et al. 2011), which is higher than that in the present study. This may be because the investments in equipment and construction, maintenance and labor costs were not included in the operating cost calculations in the present study.
|Power fee ($/m3)||0.05|
|Power fee ($/m3)||0.05|
The results of operating conditions on sulfate reduction in four systems indicate an enhanced activity of SRB by adding Fe0/Cu0. The highest sulfate reduction efficiency of 92.0% was obtained with the feed sulfate loading rate of 3,000 mg/L in the Fe/Cu + SRB system, which was significantly higher than that of SRB and Fe + SRB systems. The presence of Fe0 could increase solution pH and decrease solution toxicity through the formation of metal sulfide precipitate. The sulfate reduction of the Fe + SRB system could be further enhanced by adding Cu0. This study demonstrates that the Fe/Cu + SRB system could be a promising technology for treating wastewater containing heavy metals and sulfate.
This study was funded by the Science and Technology Project of Yellow River Conservancy Technical Institute (no. 2017KXJS017).