Mechanism of Ozone Oxidation of Polycyclic Aromatic Hydrocarbons During the Reduction of Coking Wastewater Sludge

2.1 Target PAHs and sludge source

Sixteen PAHs were selected as the representative PAHs whose structures, partitioning coefficient logK_ow and carbon normalized sorption coefficient K_oc are shown in Supporting Information Tab. S1. The deuterated surrogates were naphthene-d8, acenaphthene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12. Hexamethylbenzene was used as an internal standard for gas chromatography (GC) analyses and was obtained from Sigma–Aldrich (Gillingham, Dorset, VT). All solvents used for sample preparation and analysis (dichloromethane, hexane, acetone, and methanol) were HPLC grade from Merck (Darmstad, Germany). Deionized water was produced by a Milli-Q system (Millipore, Bedford, MA). The activated sludge was collected from the Songshan coking wastewater treatment plant (WWTP) located in Shaoguan, Guangdong province in China, having an average treatment capacity of 2000 m³· day⁻¹. The ammonia of effluent and cleaning wastewater were collected from the influent of the coking WWTP. The primary treatments included a flotation-degreasing tank and an equalization basin. The particles and tar separated in the flotation-degreasing tank were oxidized by the Fenton method, while the primary effluent was directed to the biological system. An anoxic–oxic–hydrolytic–oxic system coupled with a biological fluidized bed was applied at the biological stage, and pressurized air was injected into the biological tanks. The biological sludge from various tanks of the biological stage were mixed with a thickener, and then pumped to the dewatering unit. The concentrations of the total suspended solids (TSS) and volatile suspended solids (VSS) were about 5.7 and 4.4 g L⁻¹, respectively, and the pH value was about 7.3. The mean concentrations of regularly measured parameters (e.g., chemical oxygen demand (COD), volatile phenols, NH₄⁺-N and CN⁻) are shown in Supporting Information Tab. S2, with the average removal efficiencies ranging from 96 to 99%.

2.2 O₃ FBR

An O₃ FBR (sludge: batch addition; O₃ gas: continuous supply; UV lamp: continuous irradiation) was used for the sludge ozonation experiments, as shown in Fig. 1. The sludge sample was oxidized in a cylindrical glass FBR (diameter 85 mm, height 300 mm) with an effective volume of 5 L. A UV lamp was installed along the inner wall of the reactor, which had a power of 20 W and was enclosed in a quartz tube. Antioxidative nano aluminum reflectors were fixed on the outer wall of the FBR. The O₃ gas was produced by continuously bubbling into the reactor pure oxygen by using an ozone generator (CF–G–3–10 g, Guolin, P. R. China; maximum yield 10 g h⁻¹). The off-gas during and after ozonation was collected in a 1 L solution of 1% KI and was then titrated by 0.1 M Na₂S₂O₃ to determine the amount of unreacted O₃. The inlet true concentration of O₃ was determined by the same iodometric method. The amount of O₃ consumed during the reaction was calculated as follows:

(1)

where represents the gaseous concentration in of O₃ (mg L⁻¹) at the inlet; represents the flow rate of O₃ gas; t is the duration (min) of the sludge ozonation; and is the total mass (mg) of O₃ in the off-gas. The amount of O₃ consumed could be readily controlled by adjusting either the ozonation duration or the electrical intensity of the O₃ generator.

2.3 Experimental procedure

Before each experiment, pure oxygen was injected into the reactor to stir the coking sludge. The ozone generator was switched on when the power and the inlet flow were stable. In all oxidation experiments, the initial concentrations of the target PAHs in the coking sludge have been determined as shown in Tab. 1. The reaction temperature was fixed at about 25°C. Samples were collected in an amber glass bottle and analyzed in duplicate for each sampling time.

Five experiments were performed under different ozonation conditions. The aim of experiment 1 was to determine the optimal range of the O₃ dose by examining the solubilization characteristics of the coking sludge. The objective of the experiment 2 was to investigate the degradation efficiency of the target PAHs in coking sludge at various O₃ doses. In experiment 3, the effect of UV irradiation on PAHs removal during coking sludge ozonation was assessed with a specific ozone dose, and in the presence and the absence of a UV irradiation. In experiment 4, the effect of adding H₂O₂ on PAHs removal during the sludge ozonation was determined by changing the dosing mode of H₂O₂ (i.e., batch or continuous mode) and varying the molar ratio of H₂O₂ to O₃ (nH₂O₂/nO₃). In the continuous mode, a total of 10 mL H₂O₂ solution was evenly pumped in to the reactor throughout the sludge ozonation process. In the batch mode the same amount of H₂O₂ solution was added into the reactor at the beginning of the reaction. In both modes, the O₃ dose was set at 400 mg O₃ g⁻¹ SS with nH₂O₂/nO₃, 1:1. To examine the effect of the molar ratio of H₂O₂ to O₃, two ratios (1.5:1 and 2:1) were also used in the batch mode. The purpose of experiment 5 was to investigate the effect of the sludge pH on the PAHs removal by adjusting the pH to 5 with 40% H₂SO₄ v/v or to 9 with 40% NaOH w/w. In this experiment, the dose was maintained at about 400 mg O₃ g⁻¹ SS. All experiments were conducted in duplicate at ambient temperature (23 ± 2°C).

2.4 Analysis

PAHs were analyzed by GC/MS (Agilent, 7890A-5973C) in the selected ion mode with a 30 m × 0.25 mm id × 0.25 μm film thickness HP-5MS fused silica capillary column. The GC/MS conditions for sample analysis were as follows: the injection port, interface line and ion source temperature were maintained at 280, 290, and 250°C, respectively. The column temperature was programmed from 60 to 310°C at 5°C · min⁻¹ and held for 10 min. Helium was the carrier gas at a flow rate of 1.2 mL min⁻¹ with a linear velocity of 42.4 cm s⁻¹. The mass spectrometer was operated in the electron impact ionization mode (70 eV). The sample (1 μL) was injected with a split ratio of 10:1.

Because of the adsorption of the PAHs, the samples must first undergo a preparation before further analysis. The sludge from O₃ FBR in each series of experiment was taken for analysis. The samples were filtered through a glass filter to separate the liquid from the solid phases. The solid phase was freeze dried and kept at − 20°C before analysis. The solid sample adsorbed onto the glass filter was spiked with 20 μL of surrogate standards (80 μg mL⁻¹) and extracted with a Soxhlet apparatus with 200 mL of dichloromethane for 48 h in a water bath maintained at 46°C. Based on the contents and concentrations of the organic compounds, 1000 mL of wastewater was extracted with C18 cartridges, to which 20 μL of the surrogate standards (80 μg mL⁻¹) were added to correct the losses along the extraction process and provide the efficiency of the extraction. Activated copper was then added to the flask to remove sulfur from the extract.

The extract of the aqueous and solid phase were separately loaded onto a 1:2 alumina/silica gel glass column with 1 g anhydrous sodium sulfate overlaying the silica gel for clean-up and fractionation. A volume of 15 mL hexane was first applied to remove aliphatic hydrocarbons. The eluents containing PAHs were then collected by eluting 70 mL dichloromethane/hexane (3:7, v/v), and were concentrated to 0.5 mL under a gentle stream of pure N₂. Internal standards (5 μL, 100 μg mL⁻¹) were added to the sample prior to the GC/MS analysis.

Quantification was determined using a seven-point calibration curve established from a hexane-based internal standard for each individual PAH. The R²-values of PAH calibration curves were all >0.99. The detection limits of the method ranged from 0.06 to 16.56 mg kg⁻¹. The average recoveries for all the sludge samples were 60.1 ± 5.6% for naphthalene-d8, 74.5 ± 10.1% for acenaphthene-d10, 88.7 ± 8.8% for phenanthrene-d10, 98.3 ± 8.5% for chrysene-d12, and 96.4 ± 9.6% for perylene-d12.

COD_Cr, TOC, and pH values were analyzed according to the Chinese State Environment Protection Agency (SEPA) Standard Methods (Chinese SEPA, 2002). TOC was measured with a TOC analyzer (TOC-VCPH, Shimazdzu, Japan) after filtration through a 0.45 μm membrane. The pH value was measured with a pH meter (pHS-3C, P. R. China).

2.5 Date processing

The rate of sludge solubilization was defined as follows:

(2)

where TSS₀ and TSS represent the initial and residual (at t, min) concentrations of TSS of coking sludge (g L⁻¹), respectively, and SR represent the solubilization rate.

Based on the concentrations of each target PAH detected in the liquid and solid phases, the total concentration in the coking sludge could be calculated as follows:

(3)

where i represents an individual PAH; C_i is the total PAH concentration in sludge (ng L⁻¹); S_i is the PAH concentration in the solid phase of the sludge (ng g⁻¹ TSS); TSS is the concentration of total suspended solids (g L⁻¹); and L_i is the PAH concentration in the liquid phase of the sludge (ng L⁻¹). Because the solid phase volume was <3% of the total sludge volume, it can be neglected in the calculations. The amount of each individual PAH in the liquid and solid phases of the sludge (1 L) was calculated as follows:

(4)

(5)

where m_l and m_s are the amounts of an individual PAH (ng) in 1 L of coking sludge in the liquid and solid phases, respectively.

The degradation rates of PAHs during oxidation were affected by the concentrations of O₃ and PAHs, as well as the reaction rate constants. The oxidation kinetics of PAHs are described by Eq. 6. Oxidation in aqueous environments is effective due to O₃, direct photolysis, H₂O₂, and OH radicals. Indeed, the OH radicals in water are involved in a chain reaction that decomposes ozone, and several impurities can promote this. In addition, UV irradiation and H₂O₂ also promote the generation of OH radicals. In this study, the range of PAHs concentrations employed was relatively low, and the formation rate of the various oxidizing agents should have been constant at fixed experimental conditions. The oxidation effects in this study could thus be safely assumed to follow first order kinetics regarding the PAHs concentrations, as shown in Eq. 6, where k is an apparent rate constant incorporating the nearly constant concentration of the oxidizing agent, C_t is the concentration of the target PAHs at time t, and C₀ is the initial concentration of the target PAHs.

(6)

(7)

3.1 Sludge solubilization characteristics by O₃ FBR

Figure 2 shows the effect of the O₃ concentration on the solubilization of the coking sludge. As the concentration of O₃ increased, the solubilization rate raised almost linearly until 300 mg O₃ g⁻¹ SS, after which point it reduced. At 500 mg O₃ g⁻¹ SS, the solubilization rate reached the highest value of 68.2%. Previous studies showed that O₃ concentrations as low as 50 mg O₃ g⁻¹ SS effectively reduced the excess of sludge production 25. As the concentrations of the PAHs and other organic substances in the coking activated sludge in this study were much higher compared to published studies, higher O₃ concentrations were needed to destroy the activated sludge 26. The solubilization of the activated sludge was the key process of the sludge reduction. Figure 2 illustrates that the concentration of soluble COD increased dramatically to 879 mg mL⁻¹ at 400 mg O₃ g⁻¹ SS compared to the initial value. The amount of proteins and polysaccharides that were measured in the sample also indicated that the sludge and microbial cells were destroyed. A distinct increase of the soluble COD concentration can be observed (Fig. 2) because O₃ disrupts the membranes of the microbial cells leading to the release of intracellular substances into the liquid phase. Figure 2 also shows that the total COD in the sludge was raised to a maximum at 400 mg O₃ g⁻¹ SS, and then decreased considerably fast as the concentration of O₃ increased. At 500 mg O₃ g⁻¹ SS, the total COD concentration was reduced by about 100 mg L⁻¹, implying segmental mineralization of organic substances by O₃ oxidation. In the view of large scale applications, the O₃ concentration should be optimized for economic reasons 25, 27. Therefore, an O₃ concentration ranging from 0 to 500 mg O₃ g⁻¹ SS was tested, with a sludge solubilization rate being close to 68.2% at 500 mg O₃ g⁻¹ SS.

3.2 Effects of the ozone concentration on the reduction of the coking sludge

The changes in the amounts of PAHs distributed in the liquid and solid phases of the activated sludge at different O₃ concentrations are shown in Fig. 3 and Supporting Information Fig. S1. The results indicated that the removal rates of the target PAHs were the same in the different sludge phases, even at a concentration of 500 mg O₃ g⁻¹ SS, and that the removal efficiencies were >95%. Furthermore, the dissolution rate of fluorine (Fluo), benzo[a]anthracene (BaA), chrysene (Chr), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), dibenzo[a,h]anthracene (DBA), indeno[1,2,3-cd]pyrene (Indp), pyrene (Pyr), benzo[g,h,i]perylene (BaP) and BgP were much faster than those of naphthalene (Naph), acenaphthene (Ace), fluoranthene (Flu), phenahthrene (Phe), acenaphthylene (Acy), and anthracene (Ant) in the liquid phases until the O₃ concentration reached 300 mg O₃ g⁻¹ SS. After this point, the target PAHs were almost simultaneously degraded as the O₃ concentration increased both in the liquid and solid phases of the coking activated sludge. This phenomenon implies that the PAHs in the coking activated sludge were probably degraded through three pathways. Firstly, the PAHs were released from the solid phase to the liquid phase during the process of membranolysis; secondly, the dissolved O₃ oxidized the PAHs distributed in the liquid phase; lastly the PAHs adsorbed on the solid phase were directly oxidized by the dissolved O₃. Moreover, the chemical structures of the organic substances were significantly altered by O₃ compared to their parent compounds, leading to an increase of the biodegradability of the solution 28. The results shown in Supporting Information Fig. S2 also indicate that a significant sludge reduction appeared in the O₃ FBR during the O₃ oxidation. The TSS concentration of the coking activated sludge quickly dropped as the concentration of the dissolved O₃ increased. At the same time, the TOC concentration of the liquid increased rapidly as the O₃ concentration increased and then decreased. This observation confirmed that the organic substances released from the coking sludge could be partially mineralized if sufficient O₃ was supplied to the system. Nevertheless, the organic substances in the liquid phase could not be removed thoroughly according to the residual TOC because of their enhanced hydrophilicity after the oxidation process 29.

3.3 Degradation kinetics of PAHs

The changes in the concentration of the target PAHs in the solid and liquid phases of the coking activated sludge during sludge reduction are shown in Fig. 4 and Supporting Information Fig. S3. The target PAHs Acy, Ant, Pyr, BaP, and BgP exhibited the same removal trends in the liquid phase. The concentrations increased gradually as the O₃ concentration increased, until a concentration of 300 mg O₃ g⁻¹ SS, and then the concentration gradually decreased. The initial increase and the gradual decrease in the concentrations of the target PAHs demonstrate that the degradation by O₃ in the liquid phase was lagged behind the release from the coking sludge during the sludge reduction. All target PAHs concentrations decreased as the O₃ concentration increased in the solid phase. The dissolved O₃ may alter the chemical structures of the activated coking sludge causing the release of the sludge-bound PAHs, and the direct reaction with them. The removal efficiencies of HMW PAHs were also higher in the solid phase than the liquid phase one. The adsorption process of the target PAHs in the coking activated sludge can undergo surface adsorption or partition 30. Obviously, in the first stage, the PAHs adsorbed on the sludge granules surfaces were released into the liquid phase during the initial stage of ozonation because of the solubilization of extracellular polymeric substance. And then, the dissolved PAHs in the liquid phase degraded together with the sludge reduction 28. This infers that a low O₃ concentration may lead to a significant increase in the concentrations of PAHs in the wastewater. Therefore, the practical O₃ concentration should be carefully controlled during sludge ozonation in the O₃ FBR.

The degradation rates of PAHs during oxidation were affected by the concentrations of O₃ and PAHs, as well as the reaction rate constants. In this work, O₃ was continuously supplied to the FBR and its aqueous concentration remained almost at a steady-state (about 0.13 mg L⁻¹). In order to simplify the kinetic model, the degradations of PAHs in both the solid and liquid phases of activated coking sludge were taken into consideration, thus the apparent rate constant for the reaction of an individual PAH with O₃ in activated coking sludge (ks) could be determined. Therefore, the reactions of O₃ with the target PAHs in the sludge could be approximately expressed by the pseudo-first-order kinetics. The plot of ln(C_i/C_i,0) versus t yielded a linear curve with the slope equaling ks’ under the experimental conditions of pH 7.3, and at ambient temperature of 23 ± 2°C (Supporting Information Fig. S4). The values of ks were calculated from ks’ and are summarized in Tab. 2. It is noticed that the apparent reaction rates in the coking sludge ranged from 1.91 × 10³ to 2.62 × 10³ M⁻¹ s⁻¹. Among the target PAHs, the reaction rates of the HMW PAHs in the coking sludge were higher than those of low molecular weight PAHs. A comparison of the logK_ow values (ranging from 3.37 to 6.75) shows that HMW PAHs are more easily absorbed on the coking sludge granules, and that the initial amount of HMW PAHs absorbed on the solid phase reached a maximum of 96% (Tab. 1). The adsorption or desorption of the target PAHs on the activated coking sludge greatly restricted their degradation during ozonation. In addition, the efficiency of the PAH degradation was highly dependent on the initial concentrations of the PAHs in the coking sludge, with greater degradation observed in the sludge that contained high initial concentrations of these contaminants. A number of experimental studies have explored heterogeneous reactions of PAHs with ozone 31, 32. The results indicate that these reactions proceed via a Langmuir–Hinshelwood mechanism that involves two processes: (1) equilibrium partitioning of the gas phase oxidants between the gas phase and the surface; and (2) surface reactions between the ozone and the adsorbed molecules 33. Therefore, ozone may initially be physisorbed to the surface of the sludge, after which it is transformed to a reactive intermediate prior to reacting with the PAHs. Huber et al. 34 also showed that the adsorbed micro-pollutants on sludge granules would not be oxidized efficiently by O₃ because the transfer of gaseous O₃ from gas bubbles to the liquid phase is much faster than that of aqueous O₃ from the liquid phase to the surface of the sludge.

3.4 Effect of H₂O₂ addition and ultraviolet (UV) irradiation

In the O₃ FBR, direct O₃ oxidation and indirect oxidation by formed from the reaction of O₃ could lead to the target PAHs degradation. During advanced oxidation processes (AOPs), the combination of O₃ with H₂O₂ and UV has been extensively applied to enhance generation due to the cumulative or synergistic effects 35, 36. In this work, H₂O₂ was thus added to evaluate its contribution to the PAH degradation during the sludge ozonation by varying the molar ratio of H₂O₂ to O₃ and the H₂O₂ dosing mode.

As H₂O₂ may decompose in the liquid phase, the real-time concentration of H₂O₂ in the ozonation FBR needed to be controlled. By adding H₂O₂ either continuously, or in batches throughout the experiment, the concentration of H₂O₂ was kept constant while the O₃ concentration was fixed at 400 mg O₃ g⁻¹ SS. Figure 5 and Supporting Information Fig. S5 show that the addition of H₂O₂ in combination with ozonation could increase the removal of most of the target PAHs in the contrary of an ozonation without H₂O₂. However, it is surprising that there was no statistically significant difference in the PAHs removal between the two dosing modes of H₂O₂ (Supporting Information Fig. S6). As the increased from 0.5 to 1.0, the removal of the target PAHs both in the liquid and sludge phases were enhanced. However, the removal efficiency decreased as the H₂O₂ concentration increased. It can be noted that the concentration of the target PAHs almost remained constant as the increased from 1 to 2. There are two explanations for this trend. Firstly, many organic and inorganic substances could compete for as scavengers, such as , , and CH₃COOH in the coking activated sludge 37. Secondly, the decomposition of aqueous O₃ could be accelerated by H₂O₂, and thus weakened the direct oxidation of PAHs by molecular O₃. Therefore, an equilibrium between the strengthened indirect oxidation via and the weakened direct oxidation via molecular O₃ could be established in the combination of O₃ with H₂O₂. Coupling H₂O₂ to ozonation is a way to improve the efficiency of the process due to cumulative or synergistic effects 38. The pseudo first-order rate constants of the ozonation , photocatalysis and the O₃/H₂O₂ oxidation were calculated and the degree of synergy, S, was determined according to Eq. 8.

(8)

It was found that S was >0, ranging from 4.57 to 26.68 (Tab. 3). This result indicates that individual processes are not acting separately, but that there is a two-way synergy between them.

Table 3. Degree of synergy coupling H₂O₂ to ozonation in an O₃ FBR

PAH				S (%)	PAH				S (%)
Naph	0.0045	0.0017	0.0071	12.32	BaA	0.0040	0.00086	0.0064	24.57
Acy	0.0046	0.0012	0.0067	13.22	Chr	0.0059	0.00069	0.0069	4.57
Ace	0.0033	0.0011	0.0052	15.33	BbF	0.0048	0.00076	0.0068	18.55
Flu	0.0038	0.0013	0.0054	5.54	BkF	0.0054	0.00058	0.0069	13.21
Phe	0.0033	0.00089	0.0057	26.68	DBA	0.0052	0.00042	0.0059	5.25
Ant	0.0043	0.00076	0.0062	18.34	BaP	0.0040	0.00032	0.0055	21.33
Fluo	0.0047	0.00082	0.0060	8.63	Indp	0.0051	0.00031	0.0064	15.25
Pyr	0.0038	0.00071	0.0061	26.58	BgP	0.0054	0.00029	0.0068	16.76

Additionally, there was no significant difference in the degradation of PAHs in the presence and absence of UV irradiation, the irradiation range was from 210 to 270 nm with a dominant wavelength of 235.7 nm (Supporting Information Fig. S7). This may be the result of the black coking sludge covering the surface of the quartz tubes which gradually prevents the spread of UV light.

3.5 Effect of pH of the sludge

Since the pH value has a significant effect on the decomposition of aqueous ozone, the impact of the pH on the removal of the target PAHs was investigated at pH 4.5, 7.2, and 9.2. Figure 6 and Supporting Information Fig. S8 show that when the sludge pH was 9.2, the removal efficiencies of the target PAHs in both the liquid and sludge phases significantly increased by >25% at pH 4.5. Two possible explanations are as follows 39: (1) during ozonation of PAHs, formation was increased as the pH value increased, which increased the degradation of the PAHs; and (2) the dissociation of PAHs under alkaline conditions enhanced their degradation by O₃, because O₃ reacts with dissociated PAHs much faster than with undissociated ones. The pH of the solution significantly influences ozone decomposition in water. Ozone decomposition proceeds with the following five-step chain reaction, as is shown in Eqs. 9-13 40. These equations show that alkaline conditions can increase ozone decomposition to form free radicals such as hydroxyl radicals (), superoxide , and ozonide ions . These radicals are formed both in water and on the particle surfaces at a wide range of pH values. The target PAHs thus decomposed at an alkaline pH due to the formation of a variety of free radicals in the O₃ FBR. The pH value is thus an important factor in the practical application of this method.

(9)

(10)

(11)

(12)

(13)