Continuous treatment of flotation collector wastewater using a membrane bioreactor

Froth floatation has been used for almost 100 years for different purposes. After it was proven to be an effective ore extraction method (Houot 1982), it came to be widely used to concentrate metals from minerals (Sis & Chander 2003). Froth floatation processes, however, generate great quantities of floatation wastewater. In China, approximately 1.2–1.5 billion cubic metres of mineral processing wastewater are produced per year; most of this mineral wastewater is discharged directly into natural waters with little or no treatment (Xinping et al. 2010). Today, researchers are exploring ways to treat mineral processing wastewater given this environmental challenge.

Organic floatation reagents such as aniline aerofloat (DDA) and potassium ethyl xanthate are the typical contaminants in wastewater generated by mineral floatation processing. Organic floatation reagents induce chemical oxygen demand (COD) and deteriorate water quality (Araujo et al. 2010). As a high-efficiency mineral collector, DDA ((C₆H₅NH)₂PSSH) is used frequently in lead sulfide mineral floatation in China (Youquan & Rongshang 1994; Lotter & Bradshaw 2010). However, environmental problems have resulted from widespread DDA use, and methods are needed to remove DDA from mineral processing wastewater.

Currently, physical–chemical methods and biological methods are used to remove DDA from mineral processing wastewater (Sun et al. 2001; Xie et al. 2002; Ze-ping et al. 2006). Physical–chemical methods consume a lot of energy, have high operating costs, and are environmentally unfriendly. Conversely, biological methods generally consume less energy and are more economically and environmentally friendly, and have therefore become more widely used in wastewater treatment (Fernandez et al. 2010; Hait & Tare 2011). Although DDA biodegradation has been reported (Wei-feng et al. 2012), little work has been done to explore the continuous removal of DDA by reactors and the DDA biodegradation mechanism.

This study used a membrane reactor (MBR) to continuously remove DDA, and then examined the biodegradation mechanism of DDA. Hydraulic retention time (HRT) and influent loads were regulated to determine optimal MBR operating conditions. The research also identified the organic intermediates of DDA biodegradation and evaluated DDA biodegradation and transfer pathways. The two outcomes of this study are: a continuous biological treatment technology to treat flotation wastewater treatment, and a better understanding of the mechanisms by which activated sludge degrades DDA in flotation wastewater.

Analytical grade reagents were obtained from Chengdu Kelong Chemical Reagent Factory. Technical grade DDA (purity >95%) was obtained from Tieling Mineral Processing Reagent Factory.

The aeration tank was inoculated with activated sludge from a secondary sedimentation tank at the Lijiao municipal wastewater treatment plant, Guangzhou, China. Simulated wastewater, with 100 mg/L DDA, was continuously fed into the reactor to acclimate the activated sludge. The reactor's dissolved oxygen (DO) concentration was maintained at approximately 4 mg/L. The COD removal rate was used to evaluate system performance during the startup period. The startup period HRT was 3 h, and air to water ratios ranged from 10:1 to 12:1. After 26 days of inoculation and acclimation, effluent COD stabilized and the target COD removal rate of 80% was achieved.

During the main part of the experiment, DDA-contaminated simulated wastewater as solo carbonaceous loading was continuously fed from the storage tank into the reactor tank from the bottom of the tank and through the membrane module. The aerobically treated water then entered the effluent storage tank and was discharged.

DO was monitored using a DO meter and controlled using two air pumps. When the DO dropped below 3.7 mg/L, aeration strength was increased; when DO exceeded 4.3 mg/L, aeration strength was decreased. The HRT was controlled by regulating the effluent flow rate. To reduce the need for activated sludge disposal and keep the high activated sludge concentration, no excess sludge was discharged for the full operating time. The system operated at temperatures between 20 and 30 °C. Backwash was conducted periodically to prevent the membrane modules from blocking. During the stable operating period, the mixed liquid suspended solids of the MBR were approximately 4.2 mg/L.

To study the of DDA biodegradation mechanism, 1 L of active sludge from the MBR aeration tank was inoculated with 1 L of 200 mg/L DDA-contaminated wastewater. The activated sludge from the reactor was moved into a small MBR for aerobic biological treatment. Samples were periodically collected from the reactor to learn about the DDA shift route and its biodegradation intermediates. Biodegradation intermediates were analyzed using a gas chromatograph–mass spectrometer (GC-MS); the DDA transfer pathway was studied by analyzing the DDA concentration in the soluble microbial products (SMP) and the bound extracellular polymeric substances (EPS).

SMP and EPS were extracted using a modified heat extraction method described by Wang et al. (2009). Sludge suspension samples of 50 mL each were collected from the small MBR and centrifuged at 3,200 rpm for 30 min to separate the solids. The supernatant was considered to be the SMP. The remaining sludge pellet was then re-suspended in a 0.9% (w/w) NaCl solution, sonicated at 20 kHz for 2 min, and then centrifuged at 4,200 rpm for 20 min. The collected supernatant was regarded as loosely bound EPS (LB-EPS). The residual sludge pellet left in the tube was then re-suspended, again with 0.9% (w/w) NaCl solution, again sonicated at 20 kHz for 2 min, then heated at 80 °C for 30 min, and then centrifuged at 6,000 rpm for 30 min. The collected supernatant was considered to be the tightly bound EPS (TB-EPS). All of the centrifuged supernatants were filtered through 0.45 μm membranes.

Influent and effluent COD were measured using the standard method (State Environmental Protection Administration 2002). The DDA concentration was determined using ultraviolet spectrophotometry as follows. First, the DDA concentration sample was centrifuged at 3,200 rpm for 5 min; the supernatant was measured using an ultraviolet spectrophotometer at 230 nm. The DDA concentration was determined using the working curve based on the absorbance; the curve range was 0–300 mg/L. (Wei-feng et al. 2012). Both COD and DDA samples were collected and analyzed in triplicate.

DDA biodegradation intermediates were analyzed using a GC-MS (Agilent 7890A-5975C, USA) equipped with a quadrupole analyzer and an HP-5MS capillary column (50 mm × 2.1 mm i.d, 0.25 μm). Mass spectra were gained at an electron impact potential of 70 eV. Helium was the carrier gas, with a flow of 2.20 mL/min. The injector and detector temperatures were set at 300 °C, with an injection mode of splitless injection. Table 1 shows the temperature program. The mass range scanned was from 50 to 650 m/z. Based on fragmentation rules of organic species under electron ionization conditions, oxidant products were identified by comparing mass spectra with National Institute of Standards and Technology (NIST) library data.

These results indicate that extending the HRT to a certain point increases organic compound removal (Cheng et al. 2012). At a shorter HRT, the DDA does not completely degrade, due to insufficient contact time. Continuing to extend the HRT, however, does not speed up the COD removal rate. Extending the HRT from 2 h to 3 h increased the average COD removal rate from 79.5 to 89.9%, and reduced average COD concentration in the effluent to 16.11 mg/L. Extending the HRT from 3 h to 4 h only slightly increased the average COD removal rate, from 89.9 to 93.9%, with an average effluent COD concentration decrease from 16.11 to 9.53 mg/L.

This levelling of performance illustrates that an extended HRT is not economical nor efficient. Further, a long HRT may reduce the sludge concentrations due to microorganism digestion, if there is insufficient carbon material to sustain these organisms, reducing MBR performance. As such, the optimum HRT for COD removal was considered to be 3 h.

This pattern may be due to the fact that DDA is a toxic organic reagent and may release an iminobenzene radical or other toxic organic compound when it degrades (Helin & Wen 2003). This toxic organic compound negatively impacts the microbe and reduces COD removal. The effluent COD concentration shows that the negative influence increased as the influent DDA increased.

Pollutants are removed by activated sludge adsorption, biodegradation, and membrane filtration. Among them, activated sludge biosorption and biodegradation are primary pollutant removal approaches (Khunjar & Love 2011). Many studies have confirmed that activated sludge adsorption and biodegradation are important mechanisms contributing to pollutant removal during wastewater biological treatment (Ternes et al. 2004; Andersen et al. 2005; Lin et al. 2015). Adsorption is driven by electrostatic or non-electrostatic interaction such as hydrophobic, van der Waal forces and hydrogen bonding (Carballa et al. 2008).

A previous study confirmed that EPS and SMP are important components in active sludge and are primary materials for pollutant absorption (Khunjar & Love 2011). The pollutant was absorbed by SMP, LB-EPS, and TB-EPS, and was finally biodegraded by microorganisms. DDA concentrations in the SMP decreased first, followed by declining levels in the LB-EPS. When the biodegradation rate was lower than the transformation rate in TB-EPS, the undegraded DDA concentrated in the TB-EPS. As Figure 5 shows, the DDA of SMP decreased significantly over the testing time, and the DDA of LB-EPS decreased slightly. The DDA of TB-EPS decreased slightly and then increased. These results may be explained by the initial adsorption of DDA by the SMP; DDA was then transferred to LB-EPS and then to TB-EPS. Finally, the biosorbed DDA was degraded by the cell. A change curve confirmed these assumptions.

To learn more about the biodegradation pathway, DDA intermediates were analyzed using a GC-MS. The biodegradation samples were collected from the reactor as described in the ‘Biodegradation mechanism experiment’ section, and then centrifuged at 4,000 rpm for 10 min to separate the supernatant. The supernatant was extracted with dichloromethane, dried with anhydrous sodium sulfate, and concentrated 10 times using a vacuum rotator evaporator (RE-52A, China). After pretreatment, the extracts were analyzed with the GC-MS.

To obtain information on intermediates, we compared the peak area (directly related with compound concentration) of each identified intermediate presented in the total current chromatograms.

Figure 6 shows that an increased reaction time leads to a decrease in peak 4 (aniline), illustrating that aniline is the main primary byproduct and is subsequently involved in secondary reactions. Peak 2 (p-xylene) increases with increased running time, illustrating that p-xylene is one of the degradation products.

Figure 6(b) shows that peak 8 (diethyl phthalate), peak 9 (ethyl citrate), and peak 16 (diisooctyl phthalate) were detected in all samples. Peak 9 and peak 16 increased to a maximum level and then decreased over time. However, peak 8 first increased and then reached a stable level. As the operation time continued, new intermediates emerged, such as peak 10 (diisobutyl phthalate), peak 13 (N-phenyl-1-naphthalenamine), and peak 15 (cholestane-3-thiol,cyanate, (3a,5a)-(9CI)). Among these, peak 13, detected in the 4th, 8th and 24th hour points, increased with time. This suggests that peak 13 was one of the degradation products. Peak 10 was detected in the 8th and 24th hours; the peak area decreased over time. This phenomenon also suggests that peak 10 was an intermediate product, which could be further degraded. Other peaks, such as peak 1, peak 3, peak 5, and peak 6, were only detected at the 24 hour point, and the peak areas were small.

This study investigated DDA treatment and the biodegradation mechanism in a laboratory-scale MBR. Influent DDA concentration negatively influenced MBR performance. We found that a 3 h HRT was efficient and economical to operate the MBR. At this HRT, the DO concentration was 4.8–5.3 mg/L; influent pH was 6.5–7.0; and the influent DDA concentration was 95 ± 5 mg/L. COD removal by the MBR fluctuated slightly, with an average removal rate of 89.9%. The DDA transfer pathway in active sludge indicated that SMP absorbed much of the DDA; absorbed DDA was transferred to LB-EPS, and then finally concentrated on the TB-EPS. The DDA concentrated in the TB-EPS was ultimately degraded by the cell. GC-MS results showed that DDA degraded to aniline and dithiophosphate; the aniline was further oxidized to another compound and then was finally mineralized through the TCA cycle.

This work was supported by Education Special Funds of University Discipline Construction of Guangdong Province (Grant No. 2014KTSP022), Nature Science Foundation of Guangdong (2015A0303003) and Science and Technology Project of Guangdong (2014B020216009).