Insights into the enhanced fluoranthene degradation in citric acid coupled Fe(II)-activated sodium persulfate system Open Access

Polycyclic aromatic hydrocarbons (PAHs), which are a class of hazardous organic chemicals with two or more benzene rings, are widely present in the environment due to the incomplete combustion of coal, fuel and wood (Keyte et al. 2013). Fluoranthene (FLT), one of the typical tetracyclic PAHs with carcinogenic property and difficult degradation (Kadri et al. 2017), is on the priority list of PAHs regulated by the U.S. Environmental Protection Agency (USEPA) due to its high content and frequent detection in various countries, including China (Wang et al. 2017). Therefore, the elimination of PAHs, including FLT, from contaminated soil and groundwater has become an urgent task, and the corresponding technologies are diversely developed.

The details of materials used in this study are available in the Supplementary material.

The FLT stock solution (0.004 M) was prepared by employing 0.0202 g solid FLT with 25 mL acetone and then stored in the refrigerator at 4 °C. All FLT degradation experiments were conducted in a 250 mL cylindrical glass reactor (an inner diameter of 6.0 cm and a height of 9.0 cm) with two openings on the top for dosing and sampling in which a magnetic stirrer (600 r min⁻¹) was used to keep the homogeneity of the aqueous solution. The temperature in all batch reactions was controlled at 20 ± 0.5 °C by a low-temperature water bath (DC, Ningbo, China) and the initial solution pH was unadjusted. According to the EPA DSSTox database, FLT has a relative molecular mass of 202.25 and a solubility of 0.20–0.26 mg L⁻¹ in.water. The FLT concentration was set at 0.001 mM which corresponds to the concentration of FLT at actual PAHs-contaminated groundwater (Ogbuagu et al. 2011). 0.001 mM of FLT was prepared by adding 0.25 mL of FLT stock solution to a volumetric flask containing 1 L ultrapure water (groundwater used for actual groundwater tests) for dissolution for 8–10 hr. The addition of acetone was so small that it could not affect the reaction. First, the FLT solution was transferred into the reactor. Second, the predetermined amounts of FeSO₄·7H₂O and CA were added to the solution successively. After Fe(II) and CA were mixed homogeneously, a designed dosage of PS was added to initiate the reaction. At the predetermined time, 0.5 mL samples were taken out to a 5 mL vial pre-filled with 0.5 mL methanol to terminate the reaction. Afterwards, the samples were filtered by 0.22 μm millipore membrane for high-performance liquid chromatography (HPLC) analysis. To investigate the influence of the initial pH of the solution, the initial solution pH was adjusted by 0.10 M NaOH or 0.10 M H₂SO₄, otherwise, the initial solution pH was unadjusted.

For the experiments concerning the influence of groundwater matrixes, sodium bicarbonate, sodium chloride or humic acid was initially added into the reactor containing FLT, then other chemicals were added later as in the FLT experiments. The remaining content of FLT was determined by HPLC at the predetermined time. In the experiment exploring the effect of surfactants on FLT degradation, the surfactant was added first and other reagents were added later, then the FLT concentration was analyzed by HPLC at the predetermined reaction time. All tests were conducted at least in triplicate and the mean values were reported.

The concentration of FLT was analyzed by HPLC (LC-20AT, Shimadzu, Japan) at the detection wavelength of 235 nm. The mobile phase was a mixture of ultrapure water and methanol (10:90 (v/v)). The oven temperature was 30 °C and the injection value was 100 μL.

The intermediates during FLT degradation were analyzed by GC-MS (Agilent, G2577A, USA) equipped with a flame ionization detector (FID) and an HP-5MS capillary column (Li et al. 2019a, 2019b). Electron paramagnetic resonance (EPR) tests for active radical analyses were conducted on a Bruker EMX-8/2.7C spectrometer using DMPO or TEMP as spin-trapping compound. The instrumental conditions are described in the Supplementary material.

The residual PS concentration was monitored using the KI colorimetric method (Liang et al. 2008). The concentrations of Fe(II) and total Fe in the aqueous solution were determined by o-phenanthroline spectrophotometry (Harvey et al. 1955). The explicit analytical methods used in this work are shown in the Supplementary material.

Figure 1 shows FLT degradation in different systems, among which the molar ratio of PS/Fe(II)/CA/FLT is 60/20/5/1. In the blank test, it can be seen that up to 4.7% FLT was lost in the solution, indicating that the volatilization of FLT was well controlled in the entire experiment. FLT degradations of 6.4 and 13.6% were achieved when only Fe(II) or PS along were added, indicating no direct reaction of Fe(II) with FLT when compared with the blank result, and the PS reaction with FLT in aqueous solution was also generally slow. For the PS/Fe(II) system, FLT was degraded rapidly in the initial 5 min (50% removal), then more than 12% was removed in the next 105 min. This two-stage degradation process was also observed in another report (Gan & Ng 2012) and it could be inferred that a large number of ROS were generated immediately with enough PS and available Fe(II) initially, but concurrently leading to massive consumption of Fe(II) and larger accumulation of Fe(III) in the following time, therefore losing the activation function. It is noteworthy that FLT degradation was increased dramatically from 62 to 96.3% in 120 min reaction after adding CA into the PS/Fe(II) system. Compared with the PS/Fe(II) system, the PS/Fe(II)/CA system had a persistent degradation of FLT in the whole reaction time, proving the significant effectiveness of CA in the PS oxidative process on FLT degradation. There may be several reasons for the enhancement of the FLT degradation by CA. On the one hand, the pH of the solution was decreased from 5.34 to 4.82 (Table 2) after adding CA, which could prevent the precipitation of irons and hold the soluble iron in the solution to a certain extent. On the other hand, CA may elevate iron solubility by forming Fe(II)/Fe(III)-CA complexes, and promote Fe(II) recycling by altering the redox conditions (Liang et al. 2004).

In order to determine whether the improvement of FLT removal was due to the effect of pH reduction or the effect of Fe(II)/Fe(III)-CA complexes, the initial solution pH of the PS/Fe(II) process was adjusted to 4.82 which was in line with the unadjusted pH of the PS/Fe(II)/CA system. Although the initial pH dropped to 4.82 in the PS/Fe(II) system, FLT removal was only 6.2% higher than the unadjusted one (Figure 2). In addition, the soluble Fe and oxidant concentrations during the reaction were determined. As illustrated in Figure S1b, the total Fe in the PS/Fe(II) system was significantly lower than that in the PS/Fe(II)/CA system, which confirmed that CA could increase the concentration of soluble iron by forming Fe(II)/Fe(III)-CA complexes. Besides, the PS consumption increased with the addition of CA, which implied the generation of more ROS, thus enhancing FLT degradation.

Finally, further investigation was conducted to explore the effect of CA dosage on FLT removal. In the experiment, the doses of PS and Fe(II) were 0.06 and 0.02 mM, respectively. The consequences, which are presented in Figure 3(c), Figure S2c and Table S3, are similar to the effects of PS and Fe(II) doses on FLT degradation as described above. When CA was set as 0.02 mM, although FLT degradation was almost unchanged, k was significantly lower than that of 0.005 and 0.01 mM CA, and t_1/2 was the opposite. This was mainly owing to the following two reasons. For one thing, ROS would react with CA causing unnecessary loss, and for another thing, excessive CA could form a very stable chelate with Fe(II), which was not conducive to the catalytic reaction of Fe(II) (Rastogi et al. 2009b). k and t_1/2 did not change significantly when CA was increased from 0.005 to 0.01 mM, so the CA dosage was determined to be 0.005 mM considering the remediation cost. Therefore, in the following experiments, the doses of PS, Fe(II), CA and initial FLT concentration were set as 0.06, 0.02, 0.005 and 0.001 mM, respectively.

Hu et al. (2019) reported that the representative anions, including sulfate and nitrate, had negligible effect on p-Nitrophenol removal in the persulfate/microwave heating system, and Diao et al. (2020) had the similar finding that no significant effect of sulfate as well as nitrate on bisphenol A degradation in a heterogeneous ultrasound-enhanced sludge biochar catalyst/persulfate process. In addition, the dose of PS was only set at 0.06 mM in this experiment, so the sulfate produced by decomposition was small enough to affect FLT degradation.

Actual groundwater is abundant in natural organic matter (NOM), and Lin et al. (2017) found that the electron-rich groups in NOM can react with and ⁠. Since humic acid (HA) is an important component of NOM, it was selected as a representative of NOM to study its effect on FLT degradation in the PS/Fe(II)/CA system. Generally, the concentration of HA substances in groundwater varies from 1 to 50 mg L⁻¹ (Li et al. 2019a, 2019b). Considering the limited chemical dosing in the experiment, the HA concentration range was set to 0.1–10 mg L⁻¹, and different concentrations were set in the range of 0.1–1 mg L⁻¹ to carefully examine the effect of HA on the reaction. As shown in Figure 4(c), it is obvious that FLT degradation decreased with the increasing HA concentration, proving HA had a significant inhibitory effect on FLT degradation. On the one hand, as mentioned above, HA could consume and and compete with FLT, resulting in a lower FLT removal. On the other hand, while HA competed for and ⁠, HA also inhibited the regeneration of Fe(II) (Wang & Lemley 2004), which led to the reduction of ROS and FLT degradation.

According to other research reports (Yu et al. 2018), several ROS (⁠⁠, and ⁠) are generated in a PS system activated by Fe(II), which may be produced in the PS/Fe(II)/CA process as well. Nitrobenzene (NB) as probe compound, anisole (AN) as both and probe compound and carbon tetrachloride (CT) as probe compound to confirm the generated ROS in this process (see Supplementary material). As shown in Figure S3, all three probes were degraded to varying degrees and especially in the PS/Fe(II)/CA system, demonstrating the presence of ⁠, ⁠, ⁠, and the addition of CA could promote the generation of ROS. As far as we know, there was no suitable ¹O₂ probe due to the insolubility of some potential chemicals in water and insensitivity to ¹O₂ (Krasnovsky et al. 2008; Entradas et al. 2020). Therefore, EPR was used to determine whether there was ¹O₂ during FLT oxidation in the PS/Fe(II)/CA system.

To further confirm the presence of ROS, EPR detection was used in this study. As shown in Figure 5, DMPO-OH adducts, DMPO-SO₄ adducts and TEMP-¹O₂ adducts were found in the PS/Fe(II)/CA system, indicating the presence of ⁠, and ¹O₂. However, probably due to the low concentration and instability of ⁠, no peaks were observed for the adducts composed of DMPO and (Fang et al. 2013).

Isopropanol (IPA) acted as a scavenger for and ⁠, furfuryl alcohol (FFA) acted as a scavenger for and ¹O₂, while at the same time TBA and chloroform (CF) acted as scavengers for and ⁠, respectively (see Supplementary material). As shown in Figure 6, FLT could be degraded by 96.3% without scavengers, while FLT degradation drastically dropped to 50.6% after the addition of TBA, which revealed that was the dominant ROS in FLT degradation. Besides, FLT degradation was also significantly inhibited after the excess IPA addition, indicating that FLT removal also depended on ⁠. Notably, 27.9% of FLT was held back with CF addition, demonstrating that a certain amount of was generated and made a non-negligible contribution to FLT degradation. Finally, FLT removal was decreased by 32.3% after FFA addition compared to the inhibition of TBA, which elucidated that ¹O₂ also contributed significantly to FLT degradation. In summary, both and were dominant ROS for FLT removal, along which performed a more important role in FLT degradation. This was mainly due to a large amount of being converted from (Equation (10)).

The intermediates of FLT degradation were detected by GC-MS, which are shown in Table S6. Several chromatograms of intermediates are supplied in Figure S4. Based on the degradation products, the following possible degradation pathway was proposed (Figure 7). First, FLT was decomposed into o-dimethylbenzene (o-xylene) and naphthalene after being attacked by ROS (the attack position is shown by the arrow). Second, attacked naphthalene and made naphthalene undergo hydroxylation to form 1,4-naphthol (Onwudili & Williams 2007), which in turn formed 1,4-naphthoquinone under continued attack by (Bunce et al. 1997). After 1,4-naphthoquinone was produced from hydroxylation, it formed phthalic anhydride by ring-opening, which generated 1,4-benzoquinone through hydroxylation successively (Gu et al. 2018). In addition, the benzene ring of o-xylene was attacked by to form two dimethylphenol isomers, which were subsequently decomposed into carbon dioxide and water by ring-opening and completely mineralization (Xue et al. 2018).

PAHs, including FLT, are commonly found in contaminated soil due to their low solubility in water and their high adsorption capacity to soil (Galiulin et al. 2002). Surfactants could act as solubilizers by reducing surface tension, which could desorb contaminants from the soil and transport them to the aqueous phase for further treatment (Lamichhane et al. 2017). However, some studies have shown that surfactants may affect the degradation of pollutants by advanced oxidation (Ying 2006). The effects of Tween 80 (TW-80), sodium dodecyl sulfate (SDS), polyoxyethylene octyl phenyl ether (TX-100) and polyoxyethylene lauryl ether (Brij-35) on FLT removal in PS/Fe(II)/CA system were investigated and their properties are listed in Table S2. The surfactant concentrations in actual contaminated site remediation are generally from 1.0 to 10 g L⁻¹ (Besha et al. 2018), and the concentration was set as 1.0 g L⁻¹, considering the limited dose applied in this system.

As shown in Figure 8, both k and FLT degradation showed a significant decrease after the addition of surfactants, which proved that surfactants could inhibit FLT removal in this system. For one thing, surfactant, as an organic compound that could be decomposed by oxidation, would compete with FLT for ROS (Mousset et al. 2014). Sun et al. (2021) found that TW-80, Brij-35 and TX-100 could inhibit trichloroethene (TCE) degradation by competing with TCE to consume through EPR experiments. For another thing, when the surfactant concentration was greater than 1.0 critical micelle concentrations (CMC), the surfactant would exist as micelles, which consist of the hydrophobic core and hydrophilic shell, and with regard to hydrophobic FLT molecule, it would tend to be stuck in the micelle core (Trellu et al. 2017). In the case of FLT molecules trapped in the micelle core, ROS need to attack the micelle and open it before they could degrade FLT. In addition, the effects on various concentrations of TW-80 and SDS, which are representatives of popular nonionic and anionic surfactants, were also investigated on FLT degradation. It is worth noting that the inhibition effect of TW-80 was slightly stronger than SDS at the same concentration, probably because the CMC of TW-80 is much smaller than SDS, so it could create more micelles in the solution thus requiring more ROS to open it.

To investigate the applicability of the PS/Fe(II)/CA system for FLT degradation, experiments were conducted with actual groundwater instead of ultrapure water. The major characteristics of the actual groundwater are displayed in Table S5. As shown in Figure 9, when the molar ratio of PS/Fe(II)/CA/FLT was 60/20/5/1, FLT removal was only 11.1% in 120 min. The significant decrease in degradation was mainly due to the adverse effects of the high actual groundwater pH and the high concentration of HCO₃⁻. The effect of higher agent doses on FLT degradation was investigated and the results were less than satisfactory. Therefore, the initial solution pH was adjusted to 4 using 0.1 M H₂SO₄, and a significant increase of FLT degradation was unexpectedly acquired. At pH = 4, FLT degradation reached 90.7 and 93.9% at 180/60/60/1 and 300/100/100/1 molar ratio of PS/Fe(II)/CA/FLT (Table 2), respectively, mainly because H⁺ eliminated the inhibitory effect of HCO₃⁻ by reacting with it under acidic conditions. Considering that the pH of the solution after the reaction were all below 4, natural alkaline substances such as lime could be added to the restored groundwater (Long et al. 2014). Clearly, adjusting the initial solution pH after an appropriate increment of chemical dosage was a feasible way to remediate FLT in groundwater.

In this research, 96.3% FLT removal was achieved in 120 min when the PS/Fe(II)/CA/FLT molar ratio was 60/20/5/1 and the % RSE in PS/Fe(II)/CA/FLT was 2.9%. Compared with the PS/Fe(II) system, the addition of CA could dramatically improve FLT degradation performance by chelating Fe(II)/Fe(III) and producing more ROS. The solution pH had a strong influence on the PS/Fe(II)/CA system, and the acidic condition was favorable for FLT removal. HCO₃⁻ and HA had a certain inhibition on FLT degradation, which became obvious as the concentrations increased, but Cl⁻ had almost no effect on FLT degradation. The results of probe tests, EPR detection and scavenging experiments indicated that and acted predominantly on FLT degradation. The possible FLT degradation pathway was proposed based on the detection of o-xylene and naphthalene during FLT degradation by GC-MS. The FLT molecule was attacked by ROS to split into o-xylene and naphthalene, where naphthalene and o-xylene eventually formed CO₂ and H₂O through hydroxylation, ring-opening, oxidation and addition reactions, ring-opening, and mineralization, respectively. The presence of surfactants inhibited FLT degradation and the inhibition was slightly pronounced for surfactants with low CMC at the same concentration. Experiments conducted in actual groundwater indicated that the PS/Fe(II)/CA system could be used to remediate FLT in actual groundwater.

This study was financially supported by a grant from the National Natural Science Foundation of China (No. 41977164).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

All relevant data are included in the paper or its Supplementary Information.