This study investigated the removal of chemical oxygen demand (COD), NH4-N, and perfluorinated compounds (PFCs) in the effluent from a wastewater treatment plant (WWTP) using ZnO coated activated carbon (ZnO/AC). Results suggested that the optimal dosage of the ZnO/AC was 0.8 g/L within 240 min of contact time, at which the maximum removal efficiency of COD was approximately 86.8%, while the removal efficiencies of PFOA and PFOS reached 86.5% and 82.1%. In comparison, the removal efficiencies of NH4-N, PFBA, and PFBS were lower, at approximately 47.9%, 44.0%, and 55.4%, respectively. In addition, COD was preferentially adsorbed before PFCs and NH4-N, when the contact time ranged from 0 to 180 min, and the order of PFCs removal showed a positive correlation with C-F chain length. The kinetic study revealed that the removal of COD, NH4-N, and PFCs could be better depicted and predicted by the Lagergren quasi-second order dynamic model with high correlation coefficients, which involved liquid membrane diffusion, intraparticle diffusion, and photocatalytic reactions. The saturated ZnO/AC was finally regenerated using ultrasound for 3 h and retained excellent performance, which proved it could be considered as an effective and alternative technology.
The ZnO-coated activated carbon was prepared via a homogeneous precipitation method.
Enhanced removal of PFCs in effluent using ZnO/AC absorbent was developed.
The optimum parameters, kinetics, and mechanism of PFCs removal were investigated.
Perfluorinated compounds (PFCs) are a class of fluorine-containing chemicals that have been used worldwide in plastic, rubber, leather, and other consumer and industrial products (Tomy et al. 2009; Fromme et al. 2010; Ansari et al. 2016). Due to the high C-F bond energy (116 kcal/mol), PFCs show thermal stability and repel both water and oil, making them generally resistant to degradation and thus persistent in the environment (Jian et al. 2017; Wang et al. 2018). The long-chain analogues (C-F chain length ≥7), such as perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) with hydrophilic/oleophobic properties, are the anionic surfactants most frequently detected in the environment (Sorlini et al. 2015; Fagbayigbo et al. 2018). Owing to their persistency, bioaccumulation and toxicity, PFCs have become a hotspot in recent studies. A study of the occurrence of PFCs in New Jersey public drinking water systems found that PFOA was the most commonly detected PFC (57% of samples) with a maximum concentration of 100 ng/L. Both PFOS and PFOA were also found in infant blood at average concentrations of 4.2 ng/mL and 2.67 ng/mL, respectively (Niu et al. 2013).
Chronic exposure to PFCs may cause hepatotoxicity, neurotoxicity, behavioral disability, and genetic defects that exhibit toxicity to human health (Seo et al. 2018). Various advanced treatment methods include, but are not limited to, adsorption, ultrasound-chemical treatment, membrane filtration, and electrochemical technologies (Moriwaki et al. 2005; Yu et al. 2014). These treatments have been extensively reported to degrade PFCs in prepared solutions (Hori et al. 2006; Niu et al. 2017). Considering factors such as processing costs, operating conditions, and convenient operation, adsorption is a promising method for the removal of PFCs from an aqueous solution (Ochoa-Herrera & Sierra-Alvarez 2008; Aydin et al. 2008; Senevirathna et al. 2010). Adsorbents such as activated carbon, zeolite, and silica gels used for the removal of PFCs from wastewater were well studied (Zhang et al. 2018; Su et al. 2019). Heterogeneous photocatalytic materials were also introduced as adsorbents because of their high efficiency in organic pollutant degradation in wastewater (Raizada et al. 2014). ZnO has been widely employed in the treatment of organic pollutants due to its optical and electronic properties (band gap = 3.37 eV), chemical stability, and non-toxicity among various semiconductor photocatalysts (Yong et al. 2016).
Most recent studies were performed on laboratory-prepared solutions, while limited data exist for the removal of PFCs in wastewater treatment plant (WWTP) effluent. Considering that a practical wastewater sample contains a variety of contaminants with mutual interference amongst them, the use of a laboratory-prepared water sample does not accurately simulate it. In this study, ZnO-coated activated carbon (ZnO/AC) was synthesized using a homogeneous precipitation method, and employed to remove PFCs in the effluent from a WWTP. The comparative removal performance of PFCs and other contaminants (i.e., chemical oxygen demand (COD) and NH4-N) from the effluent using ZnO/AC adsorbent was investigated.
MATERIALS AND METHODS
The chemicals include methanol (GR), methyl tert butyl ether (MTBE, 99%), ammonium acetate (chromatographic purity), zinc nitrate (Zn(NO3)2·6H2O, AR), carbamide (CON2H4, AR), ethanol (CH3OH, AR), glacial acetic acid (GR, >99.8%) and ammonia (GR, 50% v/v). Perfluorobutanoic acid (PFBA), PFOA, perfluorobutane sulfonate (PFBS), PFOS, perfluorodecanoic acid (PFDA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), perfluorononanoic acid (PFNA) and perfluorohexyl sulfonic acid (PFHxS) were obtained from Sigma-Aldrich company (USA). Quantitative mixed standard includes quantitative mixed standard PFAC-MXB, internal standard 13C4-PFOS, internal standard 13C4-PFOA, and internal standard 13C4-PFBA, which were purchased from Wellington Laboratories (Canada). Deionized (DI) water was prepared by the Milli-Q Advantage A 10 system (Millipore, USA) at a resistivity of 18.2 M·Ω. The filter membrane (0.45 μm, glass fiber filter) was purchased from Sartorius Stedim Biotech company. The Oasis WAX 6 cc pre-processing column was purchased from Waters company (USA).
Water sample collection
The practical wastewater used in this study was collected from the effluent of a WWTP in Miyun District, Beijing, which is near a food processing factory and an automobile repair workshop. This plant has a treatment capacity of 800 m3/d by employing a typical Anaerobic-Anoxic-Oxic (A2/O) system. The detailed operational conditions of this WWTP are: mixed liquor suspended solids (MLSS) concentration = 1,600 ± 150 g/m3, solids retention time (SRT) = 6 d and hydraulic retention time (HRT) = 5.0 h). All the wastewater samples were stored in polymeric plastic bottles at 4 °C before use. The pH of the water sample was measured to be 6.8 ± 0.2 on site with a pH meter (Orion GS9156, Thermo Fisher Scientific Inc.). The concentrations of COD, BOD5, NH4-N, TP were 92 ± 2.8, 28.5 ± 3.3, 11.2 ± 1.5, and 0.9 ± 0.2 mg/L, respectively, after being filtered with a microfiltration membrane (0.45 μm, glass fiber filter) according to the standard methods of China (Xi et al. 2008).
Synthesis and characterization of ZnO/AC
ZnO/AC adsorbent was prepared by the homogeneous precipitation method using Zn(NO3)2·6H2O and CON2H4 according to the previous report, with minor modification (Raizada et al. 2014). In a typical procedure, the commercial coal-based activated carbon was first washed with DI water until there was no obvious change in pH, then dried in an oven at 105 °C. CO(NH2)2 and Zn(NO3)2·6H2O were accurately weighed to a molar ratio of 1:5, which was previously determined to dissolve in 50 mL of DI water. After homogeneous mixing, the solution was diluted to 200 mL. Then, 10 g of activated carbon was added and the slurry was placed into a pressure steam sterilizing pot (0.2 MPa) to react for a period of 2 h under 121 °C to obtain the ZnO/AC precursor. The ZnO/AC precursor was then calcined in a tubular furnace (N2 protection) at 500 °C for 2 h to harvest the final product.
The textural characteristics of the ZnO/AC were recorded by N2 gas adsorption isotherms at −196 °C using an ASAP 2020 (Micromeritics). Before adsorption experiments, the ZnO/AC sample was degassed at 150 °C under a vacuum of 10−6 Torr (Micromeritics). The iodine value (IV) and methylene blue value (MBV) were used to characterize the pore structure of these absorbent samples by the standard method (Iodine number, China method GB/T 7702.7-2008, and methylene blue adsorption, China method GB/T 7702.6-2008). The morphologies and elemental composition of the ZnO/AC sample were examined by a Nova nano scanning electron microscope (SEM) (s4800, Hitachi Co. Ltd) and energy dispersive spectrum (EDS) (EX250, Horiba Co. Ltd, Japan).
Batch adsorption experiment
To evaluate the adsorption capacity of synthesized ZnO/AC absorbent, batch adsorptive experiments of wastewater were conducted at various conditions. For a single adsorption, a total amount of 0.2–1.2 g of absorbents was added into 1.0 L of filtered effluent sample, and then 100 mL of prepared sample solution was poured into a 250 mL conical flask respectively before they were rotated on a horizontal vibration water bath while processing. The effects of adsorbent dosage (0.2–1.2 g/L) and contact time (0–5 h) on the adsorption performance were investigated according to the removal efficiencies of PFCs. Besides, to verify the ZnO photocatalytic effect on the removal of contaminants in the wastewater, multi-group parallel experiments were carried out under solar irradiation (Site: Haidian District, Beijing, outdoors, average temperature: 34 ± 3 °C, Grade of UV Index: 4, UV radiation: 15–30 W/m2, experiment duration: 11 a.m. to 15 p.m.) and dark environment (Laboratory room, temperature: 26 ± 2 °C), respectively. Subsequently, the saturated ZnO/AC absorbent was regenerated by ultrasonic process for different times (1–4 h).
PFCs were analyzed by using high performance liquid chromatography (HPLC) with a dual pump and autosampler (Ultimate 3000, ThermoFisher Scientific), coupled to electrospray ionization tandem mass spectrometry, and a Chromeleon 6.70 chromatography workstation (American Dionex Company). The analysis column was an Acclaim 120 C18 column (4.6 × 150 mm, 5 μm, American Dionex Company). The sample volume was set to 10 μL with 1 μL/min of flow rate. The two internal standards (2 ng 13C4-PFOA and 2 ng 13C4-PFOS) were mixed with other standard solution and added to the samples before extraction. The recovery and reproducibility of PFCs in filtered samples ranged from 79% to 113%, and the coefficients of determination (r2) of calibration curves for target analytes were all higher than 0.995. The limit of detection (LOD) was defined as the concentration that yielded a signal-to-noise (S/N) ratio of 3.0, while the limit of quantification (LOQ) was defined as the concentration that yielded a S/N ratio of 10. More details were described by our previous study (Shi et al. 2010). By using the above method, the PFCs of the filtered effluent sample was analyzed and the results are summarized in Table 1. As indicated, eight PFCs were detected in the wastewater and the concentration of PFOS was approximately 175 ng/L, followed by PFOA (166 ng/L), PFBS (88 ng/L) and PFBA (76 ng/L).
RESULTS AND DISCUSSION
Physical properties and adsorption characteristics of ZnO/AC
The physical properties of the original AC and ZnO/AC are summarized in Table 2. The granularity and filling proportion changed slightly after coating with ZnO. The IV increased from 820 mg/g to 995 mg/g and the MBV decreased from 155 mg/g to 130 mg/g, suggesting that the number of micropores of activated carbon increases, which is beneficial to the adsorption of micro-pollutants in water (Zhang et al. 2018). The pore width distribution and N2 adsorption-desorption isotherms of ZnO/AC are presented in Figure 1. As shown in Figure 1(a), there was a strong peak in the range of 0.6–1.5 nm, followed by two weak peaks in the range of 3.0–6.0 nm and 6.0–12.0 nm respectively, which indicated that the ZnO/AC was mainly dominated by micropores. The N2 adsorption-desorption isotherms of ZnO/AC (Figure 1(b)) exhibited a curve of type I according to the IUPAC classification, which further proved the ZnO/AC belonged to a microporous adsorbent. Additionally, when the value of P/P0 was less than 1.0, the rapid rise of the N2 adsorption isotherm implies that the pore diameters of ZnO/AC micropores and mesopores are close to the molecular diameters of adsorbate. When the value of P/P0 ranged from 1.0 to 3.0, the N2 adsorption isotherm tended to be gentle, indicating that non-microporous multi-layer adsorption occurs.
The SEM images of the original AC and ZnO/AC are shown in Figure 2. The original AC (Figure 2(a)) exhibited an uneven porous surface, which provided the possibility for ZnO particles to be attached on the pores of AC. After being doped with ZnO and calcined at 500 °C, a white layer emerged on the surface of the ZnO/AC precursor (Figure 2(b)). The element distribution of the original AC and ZnO/AC are presented in Table 2 and Figure 2(c), from which it can be seen that the original AC contained approximately 88.61% (mass ratio) of C with 11.39% of Al. In comparison, 31.73% and 14.17% O and Zn were detected in the ZnO/AC, respectively.
Effect of absorbent dosage
The effect of dosage on the removal performance with different amounts of adsorbents after 3 h of reaction time was carried out. To determine the photocatalytic effect of ZnO, batch adsorption experiments with ZnO/AC were performed in indoor (dark) and sunlight conditions (solar), respectively. In Figure 3(a), the COD removal efficiency increased significantly when the dosage of AC increased from 0.2 to 1.2 g/L. The COD removal efficiency could reach 74.6% when the dosage of AC ranged from 1.0 g/L to 1.2 g/L, which was close to the maximum removal efficiency (76.8%) of ZnO/AC in a dark environment. By comparison, the removal efficiency of COD was up to 86.8% with light irrigation. In this case, the order of removal efficiency of COD was ZnO/AC (solar) > ZnO/AC (dark) ≈ AC. Moreover, almost the same phenomenon was observed for other contaminants. With the dosage of the ZnO/AC (solar) ranging from 0.6 to 1.2 g/L, the maximum removal efficiencies of NH4-N, PFOA, PFOS, PFBA and PFBS were 43.2 to 47.9%, 77.5 to 86.6%, 67.2 to 82.1%, 38.8 to 44% and 39.5 to 55.4% with solar light, respectively (Figure 3(b)–3(f)). According to the report from Lu & Zhou (2012), the radiation intensity of sunlight ranges from 700 to 1 000 W/m2 from 12 p.m. to 14 p.m. on a sunny summer day (Jul. to Aug.), and the intensity of UV in Haidian District of Beijing was 15–30 W/m2 (grade of UV index: 4) during the experimental process. Thus, it can be speculated that the partial ultraviolet light in sunlight promoted the photocatalytic reaction of ZnO particles to further degrade pollutants (Shibin et al. 2015). Besides, as the removal efficiencies of COD and PFCs did not increase further with the increase in the amount of absorbent (>0.8 g/L) added, the optimal absorbent dosage was determined to be 0.8 g/L for the following optimization experiment.
Effect of contact duration
To evaluate the effect of contact duration on the removal of COD, NH4-N and PFCs, the adsorption test was conducted with agitation in a laboratory (dark) and outdoor environment (solar), respectively for a period of 300 min. As depicted in Figure 4(a), the removal efficiency of COD increased rapidly to approximately 60% in 0–80 minutes, and reached the maximum value of 70.5% after 180 minutes by using AC absorbent, while the COD removal efficiency was 76.8% and 85.2% using ZnO/AC in the absence and presence of solar light, respectively. On the other hand, these absorbents had a poor removal capacity for NH4-N as their removal efficiencies only reached 41.2, 43.8 and 47.9% using AC, ZnO/AC (dark) and ZnO/AC (solar) after 240 min of contact time (Figure 4(b)). Figure 4(c)–4(f) present the removal performance of PFOA, PFOS, PFBA and PFBS. Obviously, the removal efficiencies of PFOA (68.2–86.5%) and PFOS (67.0–82.1%) were higher than that of PFBS (41.0–55.4%) and PFBA (37.5–44.0%) in all experiments, indicating the adsorption capacity of PFCs was AC < ZnO/AC (dark) < ZnO/AC (solar) as well. For the difference of COD and PFCs removal efficiencies by adsorbent in sunlight (solar) and indoor (dark) environments, analysis of variance (ANOVA) was employed to reveal the contribution of sunlight to the effect of adsorbent application. As expected, it was demonstrated that the ZnO catalytic material had a significant contribution to the removal of COD and PFCs under sunlight, for their p-values were calculated as 0.0096 and 0.032, respectively. Notably, the removal efficiencies of PFBA (44.0%) and PFBS (55.4%) were much lower than that of PFOA and PFOS after 240 min of contact durations, which could be attributed to their strong hydrophilic group, and a shorter carbon chain made them less susceptible to degradation (Niu et al. 2013). Based on the above results, the optimal contact time was determined to be 240 min with the absorbent dosage of 0.8 g/L.
Competitive adsorption and kinetics
Since the organic matter, inorganic compounds, and NH4-N existing in the wastewater might disturb the adsorption efficiency of PFCs, herein the competitive adsorption process was studied on the ZnO/AC (solar) at different contact durations. Furthermore, to further examine the competition of different kinds of perfluorocarbons, six perfluorocarbons (i.e. PFOA, PFOS, PFHxA, PFNA, PFBA, and PFBS) with different C-F chain lengths were also analyzed. As displayed in Figure 5(b), the removal efficiencies of PFOA and PFOS were higher than those PFCs with short C-F chains, which could be ascribed to the difference in the distribution coefficient value (Kd) of the PFCS (Arvaniti et al. 2012). Despite the higher initial concentration of PFOA and PFOS possessed, the carbon-rich substances of PFCs were more preferentially adsorbed on the adsorbents than the short C-F chain ones under the competitive adsorption process, which was consistent with the previous result (Maneerung et al. 2016).
As presented in Figure 6(c), the fitted model of COD and PFCs could be split into two linear relationships (phase 1 and 2), suggesting that the adsorption process was affected by the intraparticle diffusion and the liquid film diffusion. Meanwhile, the adsorption process of COD showed a good linear relationship in the first two hours, indicating that the organic matter was adsorbed on the surface of ZnO/AC after passing through the liquid film in this period. Furthermore, most of the pollutants simultaneously underwent a diffusion process within the particles while the pollutants were accumulated to a certain extent that led to the adsorption rate being controlled by both processes.
Previous studies have demonstrated that the adsorption of pollutants in aqueous solutions by activated carbon depends on numerous factors such as porosity, microstructure, and the physical-chemical properties of activated carbon (Maneerung et al. 2016). Because the high bond energy of C-F in PFCs and the low Van der Waals force of the fluorine atom caused a strong polarity of C-F bond, the reaction between negatively charged nucleophiles and carbon atoms could be blocked. However, the doping of ZnO could change the ratio of micropores and mesopores, and provided more positive charge to increase the reactive sites on the activated carbon surface (Ochoa-Herrera & Sierra-Alvarez 2008). Additionally, due to the photoelectric property of ZnO, it is believed that photodegradation processes were involved in the removal of COD and PFCs. As depicted in Figure 7, when ZnO/AC was irradiated with solar light, charge separation produced an electron-hole pair (h+vb/e−cb). During the photosensitization process, the AC was excited to a higher energy state and transferred its electron to the conduction band of ZnO. The conduction band electrons were then transferred to the ZnO/AC surface before the conduction band electrons generated •OH radicals by reacting with O2 (Raizada et al. 2014). The valence band holes reacted with H+/H2O at the adsorption catalyst surface to produce •OH equally (He et al. 2018). As a result, the •OH in the solution could oxidize and degrade some organic matter that contributed to the elimination of toxic pollutants.
Regeneration of saturated ZnO/AC
The ultrasonic method was used for the regeneration of ZnO/AC owing to its convenience and the high efficiency at which the mechanical and cavitation effect of an ultrasonic wave can effectively promote the regeneration of activated carbon (Yu et al. 2014). In this work, the saturated ZnO/AC was regenerated by ultrasonic method and the removal efficiency of PFCs was used to evaluate the adsorption performance of ZnO/AC after regeneration. Multiple experimental results indicated the PFCs removal efficiencies could reach 55.8, 63.4, 68.5, and 64.9% after 1, 2, 3, and 4 h of ultrasonic durations, which suggesting that 3 h of ultrasonic duration could effectively remove the adsorbed pollutants in the ZnO/AC pores and achieved a regeneration state. The SEM images (Figure 8) reveal that the morphological and microstructure characteristics of ZnO/AC after used was wrapped up by pollutants, while obvious pores could be seen on the surface after regeneration, which exhibited good regeneration ability.
In this study, ZnO/AC was prepared using the homogeneous precipitation method and was used to remove contaminants in a wastewater treatment plant effluent. The results indicated that the ZnO/AC absorbent had a significant effect on the removal of pollutants under sunlight. The maximum removal efficiencies of the COD, PFOA, and PFOS reached 86.6%, 86.8%, and 82.1%, while the removal efficiencies of NH4-N, PFBA, and PFBS were less than 56% at the optimal parameters of absorbent dosage 0.8 g/L and contact time of 240 min. Moreover, there was a competitive adsorption process between COD and PFCs in the following order: COD > PFOA > PFOS > PFNA > PFBS > PFBA. The study of the adsorption kinetics illustrated that the removal of contaminants in the effluent by ZnO/AC followed a Lagergren quasi-second order model. The effective elimination of PFCs in this study was conducive to facilitating the further implication of toxicity reduction and indicated the significant potential of ZnO/AC absorbent in practical wastewater treatment.
This work was supported by the National Key R & D Program of China (No. 2018YFC0406400).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.