Abstract
In this study, the transformation and degradation mechanisms of refractory organic matter in biologically treated leachate from a semi-aerobic aged refuse biofilter (SAARB) in a nano-Fe3O4 enhanced ozonation process (nFe3O4-O3) were investigated in batch experiments. A continuous experiment then confirmed the effectiveness of the process for SAARB effluent treatment. In a batch experiment, the effects of influencing factors, including nFe3O4 dosage, O3 dosage and initial pH on the treatment performance of nFe3O4-O3 process, were comprehensively investigated. The results showed that when the nFe3O4 dosage = 6 g L−1, O3 dosage = 0.15 L minute−1 and initial pH = 7, the total organic carbon, absorbance at 254 nm and colour number removal efficiencies were 40.58%, 62.55% and 89.80%, respectively. In addition, most of the humic- and fulvic-like substances in the SAARB effluent were removed, and the condensation degree, aromaticity and humification degree of the organics were substantially reduced. The morphology and elemental valence state analysis showed that the nFe3O4 in the process was relatively stable and could form an nFe3O4-organic complex. Therefore, the probability of organics reacting with hydroxyl radical increased and the oxidation efficiency was enhanced. In the continuous experiment, both the O3 dosage and hydraulic retention time (HRT) were the key influencing factors. The treatment efficiency of the nFe3O4-O3 process was enhanced at a higher O3 dosage and longer HRT. The electrical energy consumption of the continuous nFe3O3-O3 process was calculated to be 17.72 kW h m−3 in SAARB effluent treatment. This study proved the feasibility of biologically treated landfill leachate treatment by the nFe3O3-O3 process.
Keywords
Graphical abstract
Introduction
With the rapid growth of the global population and the speed of urbanisation, the production of municipal solid waste (MSW) is also increasing. According to the China Statistical Yearbook (China NBoSo, 2020), the amount of MSW disposed off in China was 24,206.2 tons in 2019, with sanitary landfill being the main MSW disposal method. During the sanitary landfilling of MSW, a large amount of landfill leachate is generated due to the percolation of rainfall, decomposition of MSW and degradation of organics (Ehrig, 1983; Gu et al., 2020b). If the landfill leachate is not properly treated, it may present a pollution risk to the ecological environment around landfills and pose a threat to the health of the surrounding residents (Butt et al., 2014; Gu et al., 2022; Negi et al., 2020). Therefore, the proper treatment of landfill leachate is of great significance to their stable operation and the surrounding ecological environment.
The most commonly used landfill leachate treatment process is the combination of ‘pretreatment + biological treatment + advanced treatment’ (De Almeida et al., 2020a, 2020b; Shao et al., 2021; Wang et al., 2020a), among which the semi-aerobic aged refuse biofilter (SAARB) uses the microorganisms in the aged refuse from landfills to degrade organic matter in the leachate (Chen et al., 2019b, 2020c; Han et al., 2019). This is a very efficient and inexpensive means of landfill leachate pretreatment (Pan et al., 2020). Therefore, SAARB can be applied as a biological step in a leachate treatment chain. Although the SAARB is very effective for the treatment of landfill leachate, the SAARB effluent still contains a high concentration of organic matter, with a chemical oxygen demand (COD) of 700–1500 mg L−1 and poor biodegradability. Therefore, an advanced treatment should be further applied to treat SAARB effluent (Gu et al., 2020a; Guo et al., 2020).
Advanced oxidation process (AOP) can achieve a profound degradation for organic matters, and two of which that are practically applied for wastewaters treatment in China are O3-based AOP and Fenton process (Wu et al., 2021). The conventional Fenton process is efficient for landfill leachate treatment; however, the production of iron-based sludge leads to the secondary pollution and is considered as a kind of hazardous waste which needs to be treated properly. Many developed Fenton-like processes, like novel materials, cannot be directly applied to practical engineering due to high cost. An ozone (O3)-based AOP is a promising technology that has been widely applied to the advanced treatment of many organic pollutants and various wastewaters (Liu et al., 2021; Rekhate and Srivastava, 2020; Taoufik et al., 2021). The reaction mechanism of an O3-based AOP mainly includes direct and indirect reactions (Von Sonntag and Von Gunten, 2012). Direct ozonation can remove organic matter or groups containing unsaturated moieties from landfill leachate. In addition, the oxidation of O3 is selective, with the reaction with unsaturated moieties being much faster (Chu and Ma, 2000; Mathon et al., 2021). It is therefore difficult to completely mineralize the organic matter in waste leachate. Indirect oxidation refers to O3 under conditions such as high alkalinity or catalysis, generating highly oxidising hydroxyl radicals (·OH, etc.), through which macromolecular organic matter is degraded into small molecules and even mineralised into CO2 and H2O (Chen et al., 2020b). The enhancement of ozonation treatment efficiency by different heterogeneous catalysts, such as metal oxides (He et al., 2021; Huang et al., 2021; Jia et al., 2016; Tong et al., 2003), has been extensively studied. Previous studies have shown that Fe3O4 nanoparticles have a catalytic ability and adsorption capacity for organic pollutants (Ma et al., 2020; Wang et al., 2021). The magnetic Fe3O4 material has the advantage of being able to be recovered with a magnet.
In this study, a nano-Fe3O4 enhanced O3 process (nFe3O4-O3) was applied to the treatment of SAARB effluent for degrading and removing refractory organic matter, and the reaction mechanisms of the SAARB effluent treatment process were also investigated. The objectives of this study were to (1) compare the catalytic effect of various iron-based minerals and compare their treatment efficiency in controlled experiments; (2) investigate the effects of key influential parameters on the treatment efficiency of the nFe3O4-O3 process in batch experiments; (3) investigate the influences of O3 dosage, hydraulic retention time (HRT) and the operation mode by conducting continuous experiments; (4) analyse the degradation characteristics of refractory organics in SAARB effluent in the nFe3O4-O3 process; and (5) study the morphology and elemental valence change of nFe3O4 materials before and after reactions. This study proved the feasibility of a biologically treated landfill leachate treatment using a catalytic ozonation process and provided a beneficial reference for biologically treated landfill leachate treatment.
Materials and methods
Leachate samples
The landfill leachate used in this study was collected from a large-scale anaerobic landfill in southwest China. The landfill had been operated for 26 years, and it was closed in 2017. The daily MSW treatment capacity was 6000 tons per day. The leachate treatment capacity was ~2000 tons per day. The landfill leachate was used as the influent in the SAARB process, and the effluent from SAARB had a light brownish colour and was referred to as SAARB effluent. The SAARB process as the biological process has been widely applied to the treatment of landfill leachate; however, the effluent from SAARB as the biologically treated leachate should be further treated. The basic qualities of SAARB effluent were as follows: pH = 7.98, absorbance at 254 nm (UV254) = 4.03, colour number (CN) = 0.192, total organic carbon (TOC) concentration = 287.78 mg L−1, and specific UV absorbance at 254 nm (SUVA254) = 1.40 L (mg m)−1. According to its basic water quality parameters, the SAARB effluent was a typical refractory organic wastewater.
Chemicals
The concentrated sulfuric acid (H2SO4), sodium hydroxide (NaOH) and other chemicals used in the experiments were all analytical grade and were purchased from CHRONE chemicals (Chengdu, China). The nFe3O4 was also analytical grade and was purchased from Aladdin Reagent Co., Ltd (Shanghai, China). Nano-Fe3O4 is a black powder with magnetic properties, a specific surface area of 78 m2 g−1, an average particle size of 30 nm and a melting point of 1538°C.
Experimental equipment
Two professionally customised O3 reactors (effective volume of 0.5 L for the batch experiment and 2.0 L for the continuous experiment) were used in this study. As shown in Figure 1, the experimental setup included a pure oxygen cylinder, a flowmeter, an O3 generator (3S-T3; Tonglin Technology Co., Ltd., Beijing, China), two reactors (volume of 0.5 and 2.0 L), a magnetic stirrer, a flow pump, and saturated KI solution. The 0.5-L reactor was made of plexiglass, with an inner diameter of 5.5 cm and a height of 35.0 cm.

Schematic illustration of the batch and continuous experiments of nFe3O4-O3 process.
Experimental procedure
Batch experiment
Initially, 0.5 L of pH-adjusted SAARB effluent was transferred to the batch reactors. Then, a preset nFe3O4 dosage was added to the reactor and the O3 generator was turned on. The O3 dosage was controlled by adjusting the oxygen flow through the flowmeter. At the same time, a magnetic stirrer on the bottom of the reactor was turned on to ensure the system was fully mixed. During the reaction, 15-mL samples were taken every 5 minutes. The effect of sampling on volume loss did not significantly affect the treatment efficiency. The pH of the samples was adjusted to 9.0 using H2SO4 and NaOH, and then the samples were filtered through a 0.45-μm glass fibre membrane. The nFe3O4 materials were collected and air-dried for further characterisation. The batch experiment was repeated 3 times and the results shown in each figure were the average value. The standard deviations were used as error bars in each figure.
Continuous experiment
The O3 dosage in the continuous experiment was also controlled by the influent oxygen flow rate. The oxygen flow rates of 0.05, 0.10, 0.20 and 0.30 corresponded to O3 dosages of 4.90, 9.80, 18.92 and 29.40 mg minute−1, respectively. The reaction time was set to 120 minutes, and samples were taken at 0, 10, 20, 30, 60, 90 and 120 minutes.
The HRT (0.5, 1.0, 1.5 and 2.0 hours) was controlled by adjusting the flow rate of SAARB effluent. The reaction time was set to 150 minutes, and samples were taken at 0, 30, 60, 90, 120 and 150 minutes. All samples were pretreated following the same procedures described in section ‘Batch experiment’ before further determination.
Analytical methods
Water quality indicators were measured using standard methods (APHA, 2018). The pH was determined by the glass electrode method and TOC was determined by a TOC analyzer (TOC-L CPH CN200; Shimadzu, Kyoto, Japan). The relative aromatic compound content of the SAARB effluent was characterised by UV254 absorbance (Gu et al., 2019a). Chromaticity was expressed by the CN (Gu et al., 2018), which was calculated as follows
Samples were diluted to a certain level with secondary reverse osmosis ultrapure water and then analysed by a spectrofluorometer (Aqualog-UV-800C; HORIBA, Kyoto, Japan). The fixed excitation wavelength and emission wavelength were both 200–550 nm, with a scanning interval of 5 nm and a scanning speed of 2400 nm minute−1. The pseudo first-order constant k was calculated by −ln(Ct/C0) = −kt, where Ct and C0 are the concentration of UV254 or CN at time t or 0 minute, respectively, and k is the reaction constant (minute−1). The SUVA254 was calculated according to equation (2)
Before and after the reaction, CuKα was used as the radiation source for the materials. The tube current was 40 mA and tube voltage was 30 kV, with a scanning range of 10°–80° and a scanning mode of θ/2θ for continuous scanning. The characterisation was conducted using scanning electron microscopy (SEM: 5900LV; Jeol, Tokyo, Japan), X-ray diffraction (XRD: XD-2; Puxi, Beijing, China), energy dispersive spectroscopy (EDS: Evo18; ZEISS Jena, Germany) and X-ray photoelectron spectroscopy (XPS; Thermo Fisher Scientific, Waltham, MA, USA). Structural information regarding nFe3O4 was obtained by Fourier transform infrared spectroscopy (FTIR: TENSOR27; Bruker Co., Germany) using KBr with a wavelength range of 4000–500 cm−1.
Results and discussion
Batch experiments
Comparison of different iron-based catalysts
The catalysts used in the enhanced ozonation process were nFe3O4, pyrite (FeS2 as the main component) and magnetite (Fe3O4 as the main component). The treatment efficiencies of the enhanced ozonation processes using the three selected iron-based catalysts were compared at the same catalyst dosage of 4 g L−1. As shown in Figure 2(a) and (b), in these catalytic ozonation processes, the three catalysts enhanced organic removal to different degrees. The UV254 and CN removal efficiencies of the three catalytic ozonation processes were higher than those in the ozonation process alone, with the enhancement in treatment efficiency following the order of nFe3O4 > magnetite > pyrite. This was likely because nFe3O4 is a nanoscale material, which has a large surface area and large contact area with O3 and leachate. Therefore, the most significant enhancement was achieved using nFe3O4.

Comparison of the (a) UV254 and (b) CN removal efficiencies in the single O3, pyrite-O3, magnetite-O3 and nFe3O4-O3 processes, and the (c-d) UV254, (e-f) CN and (g-h) TOC removal efficiencies and rates in the single O3, single nFe3O4 and nFe3O4-O3 processes. Reaction conditions: nFe3O4 dosage = 4 g L−1, O3 dosage = 0.15 L minute−1 and initial pH = 7.98.
Analysis of the synergistic effect
To investigate the synergistic effect of nFe3O4 and O3 in the process, single nFe3O4, single O3 and nFe3O4-O3 processes were compared in terms of their UV254, CN and TOC removal efficiencies as well as the reaction rates when operating conditions varied. The results are shown in Figure 2(c) to (h). In the single nFe3O4 process, when the reaction time was 5 minutes, the UV254, CN and TOC removal efficiencies reached 19.91%, 34.24% and 6.12%, with reaction rates of 0.011, 0.034 and 0.006 minute−1, respectively. When the reaction time was extended to 30 minutes, the organic matter removal efficiency of the single Fe3O4 process did not change significantly. In the single nFe3O4 process, the removal of organic matter was mainly achieved by the adsorption of nFe3O4 due to the lack of oxidant. Compared with the single nFe3O4 process, the organic matter removal efficiency in the SAARB effluent was significantly higher for the single O3 process. At a reaction time of 30 minutes, the UV254, CN and TOC removal efficiencies in the single O3 process reached 54.96%, 82.52% and 35.71% with reaction rates of 0.032, 0.074 and 0.034 minute−1, respectively. Furthermore, when nFe3O4 was introduced into the O3 system, the nFe3O4-O3 process was more effective in the removal of organic matter compared to the single O3 process. At a reaction time of 30 minutes, the UV254, CN and TOC removal efficiency in the nFe3O4-O3 process was increased by 7.59%, 7.28% and 4.87%, respectively, compared to the single O3 process, and the reaction rates were increased by 0.012, 0.021 and 0.011 minute−1, respectively. This was due to the fact that in the enhanced process, O3 could directly react with unsaturated moieties of organics in SAARB effluent via an addition reaction. On the other hand, O3 was catalysed by nFe3O4 to generate a large amount of strongly oxidising ·OH, thus achieving the non-selective oxidation of organic matter and even complete mineralisation of organic matter (Chen and Wang, 2019; Wang and Bai, 2017). Overall, the nFe3O4 and O3 in this enhanced process had a strong synergistic effect, thus achieving the efficient and rapid removal of organic matter from the SAARB effluent.
Key influencing factors
The nFe3O4 dosage
In the catalytic ozonation process, the catalyst dosage is an important factor affecting the treatment efficiency of the process. A catalyst dosage that is too low may not achieve the enhancement effect (Chen et al., 2021), while a dosage that is too high will greatly increase the treatment cost. Therefore, the effect of nFe3O4 on treatment efficiency was investigated, and the results are shown in Figure 3(a) and (b). At a reaction time of 30 minutes, when the nFe3O4 dosage increased from 0 to 6 g L−1, the treatment efficiency was substantially increased. The UV254 and CN removal efficiencies were highest when the nFe3O4 dosage was 6 g L−1, with removal efficiencies of 62.55% and 89.80%, respectively. When the nFe3O4 dosage was further increased to 8 g L−1, the increase in the treatment efficiency was not significant, and the UV254 and CN removal efficiencies increased only slightly from 0.033 and 0.104 minute−1 to 0.037 and 0.109 minute−1, respectively. This could be explained by the higher nFe3O4 dosage providing more active sites; thus, catalysing O3 more efficiently to produce reactive oxygen species (ROS) (Wang and Bai, 2017). However, when the nFe3O4 dosage was too high, O3 could be decomposed to O2, decreasing the oxidant utilisation. The excess nFe3O4 would also react with radicals, therefore reducing the oxidation capacity of the process. It should be noted that both the CN removal efficiencies and rates were greater than those of UV254. This was because refractory organics (e.g. humic substances) in SAARB effluent contain many chromophores, and O3 will react directly with the unsaturated bonds in the organics via a dipole cycloaddition reaction. Therefore, the refractory organics in SAARB effluent were oxidised and degraded into lower molecular-weight organics.

Effects of (a-b) nFe3O4 dosage, (c-d) O3 dosage and (e-f) initial pH on the treatment efficiency of the nFe3O4-O3 process. Reaction conditions: nFe3O4 dosage = 6 g L−1, O3 dosage = 0.15 L minute−1 and initial pH = 7.98.
The O3 dosage
The effect of O3 dosage on the removal of refractory organics from SAARB effluent in the nFe3O4-O3 process is shown in Figure 3(c) and (d). Under the conditions of a reaction time of 30 minutes and nFe3O4 dosage of 6 g L−1, when the O3 dosage was increased from 0.10 to 0.15 L minute−1, the UV254 and CN removal efficiencies in the nFe3O4-O3 process increased from 61.81% and 89.11% to 69.93% and 91.18%, respectively, and the UV254 and CN removal efficiencies increased from 0.031 and 0.072 minute−1 to 0.041 and 0.085 minutes−1, respectively. Within this O3 dosage range, a higher content of O3 in the process promoted the radical chain reaction, therefore producing more ROS, such as ·OH, and enhancing the treatment efficiency. However, when the O3 dosage was further increased from 0.15 to 0.30 L minute−1, the enhancement of organics removal in the nFe3O4-O3 process was not significant. The UV254 removal efficiency only increased by 0.39% at 30 minutes when the O3 dosage increased from 0.15 to 0.30 L minute−1. Likewise, an increase in the UV254 and CN removal efficiencies was not obvious under these conditions. This was because the excess O3 further reacted with ·OH (Von Sonntag and Von Gunten, 2012), and therefore, any enhancement by further increasing the O3 dosage was limited.
The initial pH
The initial pH has a large impact on the treatment efficiency of O3-based processes (Qadafi et al., 2020; Wang et al., 2020b). In a single ozonation process, alkaline conditions promote the production of ·OH, and a good degradation efficiency can be obtained. In the nFe3O4-O3 process, the pH affects not only the ozonation process but also the reactivity of nFe3O4. The effect of the initial pH on treatment efficiency is shown in Figure 3(e) and (f). The UV254 and CN removal efficiencies were highest when the initial pH was 7, with removal efficiencies of 65.07% and 0.052 minute−1 for UV254 and 90.57% and 0.122 minute−1 for CN. However, acidic and alkaline conditions both decreased the treatment efficiencies of the nFe3O4-O3 process. When the initial pH was 3, the UV254 and CN removal efficiencies were 64.18% and 91.04%, respectively, and the removal efficiencies were lowest when the initial pH was 11.
In addition, the effects of nFe3O4 dosage, O3 dosage and initial pH on the removal rates of UV254 and CN were summarised, and results were shown in Figure 4. It can be concluded that increasing nFe3O4 and O3 dosage within a rational range can lead to the increase of reaction constant k. However, the change of initial pH value did not change the constant k obviously comparing to the results of changing nFe3O4 and O3 dosages. It can be inferred, on the other hand, that the nFe3O4-O3 process exhibits a wide application range (from 3 to 11).

Effects of (a) nFe3O4 dosage, (b) O3 dosage and (c) initial pH on reaction constant k in nFe3O4-O3 process. Reaction conditions: nFe3O4 dosage = 6 g L−1, O3 dosage = 0.15 L minute−1 and initial pH = 7.98.
Continuous experiment
The effect of O3 dosage
The effect of O3 dosage on the removal efficiency of the nFe3O4-O3 process in the continuous experiment is shown in Figure 5(a) to (d). Within the O3 dosage range of 0.05–0.30 L minute−1, the UV254 and CN removal efficiencies both increased with reaction time. When the O3 dosage increased from 0.05 to 0.30 L minute−1 at a reaction time of 120 minutes, the UV254 and CN removal efficiencies increased by 24.22% and 16.92%, respectively, and the removal efficiencies were also increased from 0.007 to 0.016 minute−1 (UV254) and from 0.011 to 0.045 minute−1 (CN). In addition, the mineralisation efficiency (TOC removal) was also improved at higher O3 dosages. When increasing the O3 dosage from 0.05 to 0.30 L minute−1, the TOC removal efficiency increased from 15.49% to 31.10% at a reaction time of 120 minutes.

Effects of (a-d) O3 dosage and (e-h) HRT on the UV254, CN, TOC and SUVA254 removal efficiencies in the nFe3O4-O3 process.
The SUVA254 value represents the relative content of aromatic substances in leachate and is a measure of the biodegradability of leachate samples (Deng et al., 2018, 2021). In the nFe3O4-O3 process, the SUVA254 value decreased at higher O3 dosages. At higher O3 dosages, the relative content of aromatic substances in treated SAARB effluent was lower, indicating a higher probability of biodegradation.
The effect of HRT
In continuous experiments, HRT will have a large effect on the treatment efficiency because it determines the reaction/contact time of organics in the nFe3O4-O3 process. As shown in Figure 5(e) to (h), in the early stage of the reaction, the effect of HRT was limited. This was because the SAARB effluent was not fully in contact with the oxidants in the process; therefore, the removal efficiencies were basically the same. As the reaction time increased, the treatment efficiencies under different HRT conditions varied greatly. When the reaction time was 150 minutes and HRT was 2.0 hours, the UV254 and CN removal efficiencies were significantly higher than those when the HRT was 0.5 hour. This was because a higher HRT resulted in a greater contact between refractory organics and O3; therefore, the oxidation time increased and a better treatment efficiency was achieved.
A higher HRT also improved the mineralisation efficiency. When the HRT increased from 0.5 to 2.0 hours, the TOC removal efficiency increased from 4.22% to 10.80% at a reaction time of 150 minutes. In addition, the SUVA254 value was lowest at an HRT of 2.0 hours at 0.95 L (mg m)−1. Therefore, increasing the HRT in the nFe3O4-O3 process will improve the organic removal efficiency; however, if the HRT is too long, it will increase the scale of operation and the associated operating costs.
Comparison of the batch and continuous experiments
In the batch experiment, the SAARB effluent in the reactor could be ozonated during the reaction, while in the continuous experiment, the SAARB effluent was continuously flowing, and therefore, only the SAARB effluent that was present in the reactor could be treated by the nFe3O3-O3 process. At a reaction time of 30 minutes in the batch experiment, the treatment efficiency was stable. In the continuous experiment, the leachate continuously entered and flowed out of the reactor. An improved treatment performance was achieved when the reaction/contact time was met.
Under the same O3 dosage and nFe3O4 dosage, the removal efficiencies of refractory organics in the batch and continuous nFe3O4-O3 process were provided in Fig. S1. The results showed that the UV254, CN and TOC removal efficiencies in the batch experiment were slightly higher than those in the continuous mode. Comparing the parameters shown in Table S1, the O3 volume per unit consumed in the two experiments was the same when the O3 dosage was 0.3 L minute−1. Therefore, the relatively low O3 utilisation efficiency in the continuous experiment could be attributed to the continuous flow of leachate affecting the transfer rate of O3 from the gas to liquid phases, thus reducing the effectiveness of the treatment of organic matter in SAARB effluent.
The SUVA254 values of the treated SAARB effluent in the batch and continuous experiments are shown in Fig. S2. The SUVA254 value of leachate from the continuous experiment was 0.68 L (mg m)−1, which was lower than that from the batch experiment (0.85 L (mg m)−1). Therefore, the content of refractory organic matter in the SAARB effluent from the continuous experiment was lower, suggesting that the residual organics would be more easily biodegraded.
Degradation characteristics of refractory organic matter
Characteristic molecular structure analysis
To investigate the degradation characteristics of refractory organics in SAARB effluent in the nFe3O4-O3 process, UV-Vis spectra and specific light absorbance were measured. As shown in Figs. S3 and S4, the light absorbance curves were smooth and no obvious peak was observed, indicating that the SAARB effluent contained a large number of organics and the constitution was complex (Gu et al., 2019b). As shown in Fig. S3, the light absorbance curve of the nFe3O4-O3 process was significantly lower than that of the single nFe3O4 and single O3 processes, indicating that the relative content of organics and the aromaticity of the treated leachate were much lower in nFe3O4-O3-treated leachate.
Specific light absorbance can be used to indicate the characteristic molecular structure (or property) in leachate. The absorbance values of E254 and E280 represent the aromaticity of the leachate. The ratio of E300 to E400 (E300/E400) represents the molecular weight and condensation degree of organics in leachate (Chen et al., 2018). The ratio of E240 to E420 (E240/E420) represents the coagulation degree and also the molecular weight of organics (Chen et al., 2018). The integral area of absorbance from 226 to 400 nm (S226-400) was attributed to the complex coagulated and benzene ring structure of organics (Yuan et al., 2016); therefore, S226-400 indicates the relative content of aromatic substances in leachate.
The E254 and E280 values decreased with reaction time in the nFe3O4-O3 process, as shown in Fig. S4, indicating that the macromolecular organics with high aromaticity and high hydrophobicity were transformed to smaller molecule hydrophilic organics with low aromaticity. As the reaction time increased, the S226-400 gradually decreased, as shown in Table S2. The results showed that the nFe3O4-O3 process was able to degrade the benzene ring compounds in SAARB effluent.
In the continuous experiments, with an increasing O3 dosage and HRT, the E254 and E280 values both decreased (Table S3). When the HRT increased from 0 to 2 hours, the E300/E400 and E240/E420 increased from 4.3810 and 11.4616 to 6.4176 and 18.6907, respectively. The results indicated that in continuous mode, the organics in the treated SAARB effluent had a lower molecular weight and condensation degree, and the relative content of organics was decreased. Overall, the nFe3O4 enhanced ozonation process could significantly reduce the aromaticity and hydrophobicity of organics in SAARB effluent.
Fluorescent DOM characterisation
Three-dimensional excitation and emission matrix fluorescence (3D-EEM) is widely used in the analysis of aromatic substances, that is, humic- and fulvic-like substances, in leachate (Wang et al., 2019; Zhang et al., 2020). The degradation and transformation characteristics of fluorescent dissolved organic matter (DOM) in SAARB effluent in the nFe3O4-O3, single nFe3O4 and single O3 processes were studied, and the results are shown in Figure 6. The SAARB effluent produced two main fluorescence peaks: Peak A at Ex/Em = 235–255/410–450 nm, which was attributed to fulvic-like substances with a low molecular weight and high fluorescence frequency, and Peak C at Ex/Em = 310–360/370–450 nm, which was attributed to humic-like substances with a relatively stable structure.

Three-dimensional excitation and emission matrix of SAARB effluent treated in the nFe3O4-O3, single nFe3O4 and single O3 processes.
As shown in Figure 6, in the single O3, and nFe3O4-O3 processes, there was a large reduction in the fluorescence intensity of SAARB effluent, while there was a limited reduction in the single nFe3O4 process. At a reaction time of 30 minutes, the reduction efficiency of Peak A and Peak C was 86.2% and 85.1%, respectively, in the nFe3O4-O3 process. The results showed that the nFe3O4-O3 process could effectively degrade both fulvic- and humic-like substances in the SAARB effluent.
In the continuous experiment, as shown in Fig. S5, the fluorescence intensity of Peak A and Peak C decreased as the reaction time was prolonged, which was consistent with the batch experiment result. When the reaction time reached 30 minutes and was then extended to 90 minutes, fluorescence peaks (Ex/Em = 275/340 nm) with protein-like substances were observed. They could be attributed to the dissolution of particle organic matter metabolised from biological processes. At a reaction time of 120 minutes, the protein-like substances were then degraded in the nFe3O4-O3 process. The reduction efficiency in the intensity of Peak A and Peak C was 79.72% and 75.73%, respectively, when the reaction proceeded with one cycle of HRT. Therefore, the results also demonstrated that the nFe3O4-O3 process in continuous reaction mode could also effectively decrease the relative content of humic substances, and the humification degree of the SAARB effluent could be greatly reduced.
Characterisation of Fe3O4 nanoparticles
To investigate the reaction mechanism by which nFe3O4 participated in the enhanced ozonation process, the nFe3O4 particles were characterised by XRD, FTIR, SEM-EDS, and XPS analyses. The results are show in Figure 7.

Characterisation of nFe3O4 in the enhanced ozonation process: results for (a) XRD, (b) FTIR, (c) SEM-EDS and (d) XPS.
The XRD analysis
The XRD patterns of nFe3O4 before and after the reaction are shown in Figure 7(a). The nFe3O4 peaks at 2θ of 30.19°, 35.55°, 43.21°, 56.98° and 62.64° before the reaction was coincident with Fe3O4. Compared with the fresh nFe3O4 peaks, the characteristic peak positions of the used nFe3O4 did not change, and in both fresh and used nanoparticles, only the peaks attributed to Fe3O4 were detected. The results showed that the nFe3O4 in the SAARB effluent treated by the nFe3O4-O3 process had a stable structure and maintained a strong reactivity.
The FTIR analysis
The FTIR results for nFe3O4 before and after the reaction are shown in Figure 7(b). The fresh nFe3O4 did not show any obvious absorbance. After the adsorption process using nFe3O4, the absorption peaks were similar to those after the nFe3O4-O3 process. The absorption peak at 3500–3400 cm−1 resulted from the stretching and vibration of C=C in aromatic structures and -OH in alcohol and phenol compounds. The absorption peak at 1640 cm−1 was caused by the stretching of the C=O bond. The absorption peak at 1379 cm−1 was attributed to the bending vibration of -CH3, and the absorption peak located at 642 cm−1 was the vibration peak of -CH2-. The peaks observed after the single nFe3O4 process proved the adsorption effect of nFe3O4 in the SAARB effluent. In the enhanced ozonation process, the relative transmittance of peaks at 1640 and 642 cm−1 decreased, indicating that the organics adsorbed on nFe3O4 were substantially degraded. These results confirmed the surface adsorption and oxidation effects of nFe3O4 in the enhanced ozonation process.
The SEM-EDS analysis
The SEM images (Figure 7(c) and (d)) revealed that the fresh nFe3O4 had a crystallised structure and no other impurities were observed. After the reaction, large lumps were observed on the surface of the nFe3O4. The EDS analysis indicated that in the 0- to 2-keV range, the nanoparticles were rich in elemental Fe and O and contained small amounts of elemental C and Na, indicating that the large grains were mainly iron oxide enriched areas. In the 6- to 8-keV range, only elemental Fe was detected, indicating that no other substances were introduced into the unit cell. The results also showed the high reactivity of nFe3O4 in the treatment process. It should be noted that the high sulphur content on the surface of used nFe3O4 could be attributed to the adsorption of sulphur-containing organic compounds from landfill leachate.
The XPS analysis
The full-range scan of nFe3O4 before and after the reaction is shown in Fig. S7. Three main peaks were detected: C1s, O1s and Fe2p. The Fe3O4 could be expressed in the form of FeO-Fe2O3, where the stoichiometric ratio of Fe(II)/Fe(III) is 1:2. The peak differentiation of Fe2p was performed using software XPS PEAK, and the Fe(II)/Fe(III) ratio before and after the reaction was calculated. According to the peak differentiation results for the Fe2p peak, there were two main peaks with binding energies of 710.98 and 724.68 eV, corresponding to Fe2p3/2 and Fe2p1/2, respectively, and Fe was present in the active material in the divalent and trivalent forms. In addition, the peaks at 713.94, 712.00 and 710.24 eV corresponded to Fe2p, which represented trivalent iron (Fe(III)tct) with a tetrahedral structure, trivalent iron (Fe(III)oct) with an octahedral structure and divalent iron (Fe(II)oct) with an octahedral structure, respectively. The Fe(II)/Fe(III) ratio before the reaction was calculated to be 2.00, which is consistent with pure Fe3O4. After the reaction, the Fe(II)/Fe(III) ratio was 2.06, indicating that some of the Fe(II) in the Fe3O4 material used in the experiment was converted to Fe(III), but only a small percentage underwent an irreversible conversion. Therefore, the activity was better maintained during the use of Fe3O4 materials.
Energy consumption analysis
The commercial price of the nFe3O4 used in this study was 1200 CNY/t. Considering the stability of nFe3O4, the consumption of electrical energy will likely be the major operating cost of the process. The direct electrical energy consumed in the nFe3O4-O3 process was largely due to the operation of the O3 generator. Therefore, an electrical energy consumption efficiency analysis was conducted. Under the conditions of nFe3O4 = 6 g L−1, O3 dosage = 0.3 L minute−1 and HRT = 2 hours, the continuous experiment achieved the expected treatment performance, and therefore, these parameters were used to calculate the energy required for SAARB effluent treatment in the nFe3O4-O3 process. The electrical consumption (EC) was defined as the amount of electrical energy (kW h) required to reduce the organics concentration by an order of magnitude in 1 m3 of wastewater, and was calculated using equation (3)
where P is the power (kW) requirement of the electrical devices (O3 generator and pump); t is the reaction time (minute); V is the volume (L) of wastewater treated; and C0 and C are the initial and final TOC concentrations, respectively. The constant ‘60’ converts minute to hour.
The power requirements of the O3 generator and pump were 0.16 and 0.012 kW, respectively. Therefore, the total power required was 0.172 kW. The HRT was 2.0 hours, and the volume of SAARB effluent treated was 2 L. After the continuous reaction, the TOC concentration decreased from 287.8 to 198.3 mg L−1. Therefore, the EC of the nFe3O4-O3 process was calculated to be 17.72 kW h m−3. Table 1 showed the treatment capability and EC of other AOPs, including microwave (MW)/PS, MW/H2O2, O3/H2O2 and photo-Fenton processes. It can be seen that MW-participated processes have a much higher electrical energy consumption. US/O3/EC process showed an impressive COD removal efficiency, but the reaction time lasted for 5 hours. It should be noted that raw landfill leachate and biologically treated landfill leachate are much different in DOM composition. In raw landfill leachate, the organic matter could be more easily degraded, while in biologically treated landfill leachate, such as membrane bioreactor (MBR) leachate and SAARB leachate, the remained DOMs mostly comprised refractory organics which are difficult for degradation. The nFe3O4-O3 process showed a great mineralisation performance to the refractory organics in SAARB leachate in 30 minutes and showed a relative low EC. Therefore, compared with other processes that have been investigated, the nFe3O4-O3 process has the potential to save energy while efficiently removing refractory organics from SAARB effluent.
Comparison of treatment efficiency, reaction time and EC by different AOPs in leachate treatment.
AOP: Advanced oxidation process; SAARB: semi-aerobic aged refuse biofilter; COD: chemical oxygen demand; TOC: total organic carbon; EC: electrical consumption; MW: microwave; MBR: membrane bioreactor.
Conclusion
Compared to single nFe3O4 and single O3 processes, the nFe3O4-O3 process had an enhanced removal efficiency of refractory organics from SAARB effluent. In a batch experiment, under conditions of Fe3O4 dosage = 6 g L−1, O3 dosage = 0.15 L minute−1 and initial pH = 7, the TOC, UV254 and CN removal efficiencies were 40.58%, 62.55% and 89.80%, respectively. The treatment efficiency of the nFe3O4-O3 process was verified in a continuous experiment. The results showed that O3 dosage and HRT greatly affected the performance of the process. Increasing the O3 dosage within a range of 0.05–0.30 L minute−1 and HRT within a range of 0.5–2.0 g, the treatment efficiency was also strengthened. According to UV-Vis and 3D-EEM analyses, both fulvic- and humic-like substances were substantially degraded and removed. The molecular structure of humic substances and the condensation degree, aromaticity and humification degree were profoundly reduced after the nFe3O4-O3 process. According to its morphology, elemental valence state and a functional groups analysis, the nFe3O4 in the process was relatively stable and could form complexes with organics, therefore enhancing the treatment efficiency. In addition, the EC of the nFe3O4-O3 process was calculated to be 17.72 kW h m−3. Therefore, the nFe3O4-O3 process was considered to be a promising method for SAARB effluent treatment, with high efficiency and low operating costs.
Supplemental Material
sj-docx-1-wmr-10.1177_0734242X211066229. – Supplemental material for Degradation of refractory organic matter in the effluent from a semi-aerobic aged refuse biofilter-treated landfill leachate by a nano-Fe3O4 enhanced ozonation process
Supplemental material, sj-docx-1-wmr-10.1177_0734242X211066229. for Degradation of refractory organic matter in the effluent from a semi-aerobic aged refuse biofilter-treated landfill leachate by a nano-Fe3O4 enhanced ozonation process by Yuyu Huang in Waste Management & Research
Footnotes
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author disclosed receipt of the following financial support for the research, authorship and/or publication of this article: The author gratefully acknowledge the major scientific and technological projects of Sichuan Province (2019YFS0509).
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
