1 Introduction

Currently, synthetic azo dye wastewater has attracted serious attention owing to its potential toxicity to the environment and humans (Chen et al. 2020; Pan et al. 2016). Among azo dyes, tartrazine is commonly used as a food coloring agent and is recalcitrant to biodegradation processes (Zhang et al. 2019a). Advanced oxidation processes (AOPs), which can generate powerful oxidative radicals, have generally been considered alternative methods for the treatment of nonbiodegradable and recalcitrant pollutants (Xu et al. 2020a; Li et al. 2019; Zhang et al. 2019b; Guo et al. 2020; Jiang et al. 2020; Guo et al. 2021). Common oxidants that have been used include persulfate (S2O82−) (Li et al. 2018a; Li et al. 2020a), ozone (O3) (Wang and Chen 2020), and hydrogen peroxide (H2O2) (Xu et al. 2020b).

Ox, as one of the dicarboxylic acids, is a common byproduct of most organic pollutants. The photolysis of an Fe-oxalate (Fe-Ox) complex has attracted much attention for effectively degrading different recalcitrant and nonbiodegradable pollutants via Eqs. (1–6) (Huang et al. 2017). The use of heterogeneous iron catalysts are advantageous compared with homogeneous iron salts because of the low production of ferric ions and the recycling of solids (Huang et al. 2017). Many heterogeneous iron catalysts have been adopted for the photolysis of ferric-oxalate complexes, such as maghemite (γ-Fe2O3) (Dai et al. 2018), goethite (α-FeOOH) (Lan et al. 2010), magnetite (Fe3O4) (Huang et al. 2017), and hematite (α-Fe2O3) (Lan et al. 2016). Currently, a large number of studies have focused on wastewater treatment by zero-valent iron (Fe0) because Fe0 is abundant, inexpensive, and environmentally friendly (Li et al. 2018b; Pan et al. 2020a). Our previous studies reported that Ox could enhance the activation of O2 in the Fe0/O2 system, thus improving H2O2 production (Pan et al. 2020b); additionally, ethylenediaminetetraacetic acid and nitrilotriacetic acid could modify the surface characteristics of Fe0 in the absence of UV (Pan et al. 2020b; Pan et al. 2019). Therefore, the photolysis of Ox catalyzed by Fe0 might exhibit good performance for pollutant removal.

$$ {\mathrm{Fe}}^{\mathrm{II}\mathrm{I}}\left({\mathrm{C}}_2{\mathrm{O}}_4\right){\mathrm{n}}^{3-2\mathrm{n}}+\mathrm{hv}\to {\mathrm{Fe}}^{\mathrm{II}}+\left(n-1\right){\mathrm{C}}_2{\mathrm{O}}_4^{2-}+{\mathrm{C}}_2{\mathrm{O}}_4 $$
(1)
$$ {\mathrm{C}}_2{\mathrm{O}}_4^{-}+{\mathrm{O}}_2\to 2{\mathrm{C}\mathrm{O}}_2+{\mathrm{O}}_2 $$
(2)
$$ {\mathrm{O}}_2\leftrightarrow {\mathrm{HO}}_2\left(\mathrm{pH}-\mathrm{dependent}\right) $$
(3)
$$ {\mathrm{O}}_2^{-}/{\mathrm{H}\mathrm{O}}_2+{\mathrm{H}}^{+}\to {\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{O}}_2 $$
(4)
$$ {\mathrm{O}}_2^{-}/{\mathrm{H}\mathrm{O}}_2+{\mathrm{H}}^{+}+{\mathrm{Fe}}^{2+}\to {\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{O}}_2+{\mathrm{Fe}}^{3+}=\uppi {\mathrm{r}}^2 $$
(5)
$$ {\mathrm{Fe}}^{\mathrm{II}}{\left[\left({\mathrm{C}}_2{\mathrm{O}}_4\right)\mathrm{n}\right]}^{3-2\mathrm{n}}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Fe}}^{\mathrm{II}\mathrm{I}}{\left[\left({\mathrm{C}}_2{\mathrm{O}}_4\right)\mathrm{n}\right]}^{3-2\mathrm{n}}+\mathrm{OH}+{\mathrm{O}\mathrm{H}}^{-} $$
(6)

A heterogeneous photochemical Fe0/Ox system (UV/Fe0/Ox) system was investigated in the present study for tartrazine removal. The objectives of the present study are listed as follows: (1) investigate the efficiency of tartrazine removal with the UV/Fe0/Ox system; (2) explore the affecting factors, such as the Fe0 dose, Ox dose, initial pH, and initial tartrazine concentration on tartrazine removal; (3) investigate the mechanism on tartrazine removal in the UV/Fe0/Ox system; (4) investigate the effect of Ox on the modification of the Fe0 surface; (5) clarify the effect of the H2O2 addition on removing tartrazine with the UV/Fe0/Ox system; (6) explore the performance of various chelating agents and catalysts for tartrazine removal; and (7) investigate the recycling of Fe0 for tartrazine removal.

2 Materials and Methods

2.1 Chemicals

Oxalate, FeSO4, Fe2(SO4)3, H2O2 (30%), catalase (CAT), superoxide dismutase (SOD), methanol (HPLC-grade), and phenanthroline were purchased from Aladdin (China). Tartrazine was purchased from Meryer Chemical Technology Co., Ltd. Fe0, Fe3O4, FeS, and Fe2O3 were purchased from Shanghai Jinshan Co.

2.2 Tartrazine Removal Experiments

Experiments were performed in 500 mL of working solution, containing Fe0 (0–0.8 g/L), Ox (0–2 mM), and tartrazine (2–30 mg/L) and stirred by a mechanical stirrer. The experiment was initiated by placing a preheated UV lamp (GPH150T5L/5 W/254 nm from Kadind) sealed in a glass sleeve into the working solution.

2.3 Analytical Methods

Concentrations of tartrazine were detected by a UV-visible spectrophotometer (Shanghai Mapada Company, P1) at 428 nm. The H2O2 concentration was detected by the potassium titanium oxalate method at 400 nm. The concentration of total Fe was detected by inductively coupled plasma optical emission spectrometry (Agilent 720ES, USA). Ox was measured by an ion chromatograph (ICS-900, Thermo Fisher Scientific). A morphological analysis of Fe0 species was obtained by SEM (Hitachi 4700 microscope) with a combined EDX analyzer (EDX). FTIR spectroscopy was conducted by an FTIR-840OS Shimadzu spectrophotometer (Japan) from 500 to 4000/cm. Mossbauer spectroscopy of Fe0 was analyzed using the MossWin 4.0 program.

Pollutant removal (η/%) was calculated according to Eq. (7), the removal rate pollutant (k) was fitted by the pseudo-first-order rate equation (Eq. (8)).

$$ \eta =\frac{C_0-{C}_t}{0}x100\% $$
(7)
$$ \ln {C}_t/{C}_0=- kt $$
(8)

where k is the degradation rate and C0 (mg/L) and Ct (mg/L) are the concentrations of the pollutant at time 0 and at the reaction time (t), respectively.

To demonstrate the role of oxalate, the value of f was calculated via Eq. (9).

$$ f={k}_1/{k}_2 $$
(9)

where k1 and k2 are the pseudo-first-order constants for removing pollutants in the UV/Fe0/Ox and UV/Fe0 systems, respectively.

3 Results and Discussion

3.1 Tartrazine Removal with the Different Systems

Tartrazine removal, Ox removal, Fe-ion generation, and H2O2 production were compared with the different systems. As shown in Fig. 1a, 27.1% of tartrazine could be removed within 60 min by only Fe0, which, due to reactive oxygen species, could be generated via the activation of O2 by Fe0 (Pan et al. 2020b). The addition of Ox could enhance the activation of O2, thus increasing H2O2 production, while Ox could also compete with tartrazine for •OH (Pan et al. 2020b). Thus, compared with the Fe0 system, the tartrazine removal increased only slightly in the Fe0/Ox system (28.6%). Under UV irradiation alone, 32.6% of tartrazine could be removed within 60 min, indicating that the direct photolysis of tartrazine occurred, whereas 47.4% of tartrazine could be removed within 60 min when UV was combined with Fe0, which might be because the H2O2 generated by the activation of O2 could be catalyzed under UV irradiation. Thus, formic acid (HOCOH) could be generated by reacting with O2 to form H2O2 during the photolysis of Ox (Jiang et al. 2017). Furthermore, H2O2 could be converted in •OH for removing pollutants under UV irradiation. However, only 27.1% of tartrazine could be removed within 60 min in the UV/Ox system, which might be because of the competition between tartrazine and Ox for •OH. Additionally, 72.4% of tartrazine could be removed within only 20 min in the UV/Fe0/Ox system owing to the reactions outlined in Eqs. (1–6), while tartrazine could not be further removed after 20 min which might because the tartrazine degradation intermediates generated during the reaction competing with the •OH radical. Moreover, as shown in Fig. 1b, the value of k (103) for removing tartrazine with the UV/Fe0/Ox system (59.5 min−1) was much higher than with the Fe0/Ox (3.5 min−1), Fe0 (6.9 min−1), UV (5.9 min−1), UV/Fe0 (10.9 min−1), and UV/Ox (6.7 min−1) systems. Moreover, the Ox removal with the UV/Fe0/Ox system was also much faster than with the Fe0/Ox and UV/Ox systems.

Fig. 1
figure 1

(a) Removal of tartrazine, (b) removal of Ox, (c) value of k (103), (d) generation of total Fe, and (e) H2O2 generation with the different systems. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1

Full size image

The total concentration of Fe ions generated with the different systems was also detected. As depicted in Fig. 1d, the total concentrations of Fe ions generated with the UV/Fe0/Ox system (10.4 mg/L) were also much higher than those with the Fe (5.3 mg/L), Fe0/Ox (6.4 mg/L), and UV/Fe0 (5.9 mg/L) systems. The generated Fe ions could induce the photolysis reaction Eqs. (1–6), thus increasing H2O2 production. As shown in Fig. 1e, H2O2 production with the UV/Fe0/Ox system (22.8 μM) was also much higher than that with the Fe0 (0.43 μM), Fe0/Ox (7.7 μM), and UV/Fe0 (0.54 μM) systems. The presence of more H2O2 with UV/Fe0/Ox could induce more •OH generation for pollutant removal.

3.2 Effect of Various Affecting Factors

3.2.1 Ox Concentration

Ox could react with FeIII and FeII to form Fe-Ox complexes, thus inducing Eqs. (1–6), while Ox could also compete with tartrazine for •OH. Thus, the effect of the Ox concentration (0–2 mM) on the removal of tartrazine was determined. As shown in Fig. 2a, the tartrazine removal increased with an increasing Ox concentration. The tartrazine removal within 60 min increased from 47.4 to 87.2%, while the Ox dose increased from 0 to 1 mM and then decreased to 85.5% at a 2-mM Ox dose. As depicted in Fig. 2b, the value of k (103) also increased from 10.9 to 82.2 min−1, while the Ox concentration increased from 0 to 1 mM and then decreased to 78.4 min−1 at a 2-mM Ox dose. Therefore, the optimal Ox dose was 1 mM.

Fig. 2
figure 2figure 2

a Removal of tartrazine at different Ox concentrations. b Value of k (103) at different Ox concentrations. c Removal of tartrazine at different initial pH values. d Value of k (103) at different initial pH values. e Removal of tartrazine at different Fe0 doses. f Value of k (103) at different Fe0 doses. g Removal of tartrazine at different tartrazine concentrations. h Value of k (103) at different tartrazine concentrations. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1

Full size image

3.2.2 Initial pH

The initial pH could not only affect the Fe0 corrosion rate but also affect the species of the Fe-Ox complex. FeII and [FeII(C2O4)2]2− were regarded as FeII-Ox complexes at pH < 3 and pH ≥ 3, respectively. FeIII(C2O4)2]+, [FeIII(C2O4)2]−, [FeIII(C2O4)3]2−, and Fe(OH)3 were regarded as FeIII-Ox complexes at pH < 2, pH 2–6, and pH > 6, respectively (Zhou et al. 2014). Figure 2c and d present the tartrazine removal at different initial pH values from 2 to 6, and the best performance was achieved at the initial pH of 3–4, which was due to [FeIII(C2O4)2]− and [FeIII(C2O4)3]2− being the dominant FeIII-Ox complexes that contained high photoactivity at pH 3–4. The value of k at pH 2 was lower than that at pH 3 because the reaction between FeII(pH 2) and H2O2 was much lower than the reaction between [FeII(C2O4)2]2− (pH ≥ 3) and H2O2 (Zhou et al. 2014).

3.2.3 Fe0 Dose

The Fe0 dose affected the soluble iron concentration, thus influencing the Fe-Ox complex concentration. Figure 2e and f depict the tartrazine removal and value of k at different Fe0 doses (0–0.8 g/L). The tartrazine removal and the value of k that increased with the Fe0 dose were enhanced because more Fe-Ox complexes could catalyze H2O2 faster. An excessive Fe2+ concentration would also inhibit •OH, thus decreasing the removal of tartrazine (Pan et al. 2016). Therefore, the optimal Fe0 dose was 0.4 g/L.

3.2.4 Initial Tartrazine Concentration

Tartrazine concentrations were different in various dye wastewaters; thus, the effect of the initial tartrazine concentration (varying from 2 to 30 mg/L) on the removal of tartrazine was investigated. As depicted in Fig. 2g and h, the tartrazine removal and value of k increased while the initial tartrazine concentration decreased. When the initial tartrazine concentration was 2 mg/L, the removal could reach 100% within only 2 min. A total of 67.2% tartrazine could still be removed within 60 min when the initial tartrazine concentration was 30 mg/L.

3.3 Mechanism Discussion

3.3.1 Reactive Oxygen Species

To determine the main radicals for the removal of tartrazine in the UV/Fe0/Ox system, superoxide dismutase (SOD), catalase (CAT), and methanol (MA) were selected for quenching O2•−, H2O2, and •OH, respectively (Song et al. 2019). As depicted in Fig. 3a and b, when SOD, CAT, and MA were added, the tartrazine removal decreased from 81.1 to 77.6, 53.4, and 42.8% with the corresponding values of k (103) decreasing from 59.5 to 54.2, 28.8 and 23.2 min−1. The order of scavengers in suppressing tartrazine removal was MA> CAT >SOD. The inhibitory efficiency (λ) values of the three scavengers were calculated via Eq. (10), and these values were 8.9, 51.6, and 61.1% for SOD, CAT, and MA, respectively. These results demonstrated that •OH was the major radical for the removal of tartrazine, H2O2 was the major intermediate for •OH generation, and O2•− played a minor role in the removal of tartrazine.

$$ \lambda =\left(\left(k-{k}_q\right)/k\right)\ast 100\% $$
(10)
Fig. 3
figure 3

a Tartrazine removal with different scavengers in the UV/Fe0/Ox system. b Value of k for tartrazine removal with different scavengers in the UV/Fe0/Ox system. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1; MA 2 M, SOD of 3000 U L−1, and CAT of 200,000 units L−1

Full size image

where k and kq refer to the values of k in the absence and presence of the quenching agents.

3.3.2 SEM-EDX and FTIR

The SEM images of Fe0 after 60 min of reaction with the Fe0/Ox, UV/Fe0, and UV/Fe0/Ox systems are presented in Fig. 4a–c. As shown in Fig. 4b, the surface of Fe0 was much coarser with the UV/Fe0 system than with the Fe0/Ox system (Fig. 4a) and UV/Fe0/Ox system (Fig. 4c). Fe-oxide corrosion products could cover the Fe0 surface during the reaction, thus inhibiting the Fe0 corrosion rate. However, the formation of Fe0-Ox complexes could inhibit the covering of the Fe0 surface by Fe oxides, thus guaranteeing the reactivity of Fe0. Moreover, as shown in Fig. 4d and e, the O content of the reacted Fe0 with the UV/Fe0 system (29.8%) was larger than that in the Fe0/Ox (17.1%) and UV/Fe0/Ox (21.6%) systems. These results demonstrated that more corrosion products covered the Fe0 surface with the UV/Fe0 system, which would inhibit the Fe0 corrosion rate.

Fig. 4
figure 4

SEM spectra of Fe0 after 60 min of reacting with the (a) Fe0/Ox, (b) UV/Fe0, and (c) UV/Fe0/Ox systems; (d) EDX pattern, (e) elemental percentages, and (f) FTIR pattern of Fe0 after 60 min with the different systems. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1

Full size image

The FTIR spectra of Fe0 after 60 min of reaction with the Fe0/Ox, UV/Fe0, and UV/Fe0/Ox systems were also detected (Fig. 4f). The obtained signals recorded at 1049, 1395, 1631, and 3440 cm−1 corresponded to C-O, -COO-, C=O, and O-H, respectively (Chong et al. 2016., Guan et al. 2015; Repo et al. 2013; Bennet et al. 2016). The high intensity of these oxygen-containing groups when reacting with the UV/Fe0/Ox system demonstrated the formation of the Fe-Ox complex on the Fe0 surface.

3.3.3 Mossbauer Spectra Analysis

Mossbauer spectroscopy was conducted as a quantitative method to inspect the Fe species deposited on the reacted Fe0 with different systems. As depicted in Fig. 5, there were two distinguishable compositions within the structure: a prevailing sextet line referenced to Fe0 and a doublet line referenced to FeOOH (Li et al. 2020b). Additionally, FeOOH might be formed according to Eq. (11). The percentages of FeOOH were 9.3, 20.9, and 10.2% with the Fe0/Ox, UV/Fe0, and UV/Fe0/Ox systems, respectively. Thus, more corrosion products covering the reacted Fe0 with the UV/Fe0 system might inhibit the reactivity of Fe0, and these results were consistent with the SEM results.

$$ 2{\mathrm{Fe}}^{2+}+1/2\ {\mathrm{O}}_2+3\ {\mathrm{H}}_2O\to 2\ \mathrm{FeOOH}+4\ {\mathrm{H}}^{+} $$
(11)
Fig. 5
figure 5

Mossbauer spectra of Fe0 after 60 min with the (a) Fe0/Ox, (b) UV/Fe0, and (c) UV/Fe0/Ox systems. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1

Full size image

3.4 Effect of a H2O2 Addition

H2O2 was the major intermediate for •OH generation; thus, we investigated the effect of a H2O2 addition on the removal of tartrazine with both the UV/Fe0 and UV/Fe0/Ox systems. As depicted in Fig. 6a and b, the H2O2 addition could significantly enhance the tartrazine removal with both systems; this result was due to the additional H2O2 improving the generation of •OH. Fig. 6c presents the value of k with different H2O2 additions to both systems. When the H2O2 addition increased from 0 to 50, 100, 200, and 500 μM, the values of k (103) increased from 10.9 to 15.9, 22.5, 26.4, and 51.3min−1 with the UV/Fe0 system and from 59.5 to 86.5, 115.6, 167.2, and 325.9 min−1 with the UV/Fe0/Ox system, respectively. Compared with the UV/Fe0 system, the value of k could be significantly enhanced with the UV/Fe0/Ox system, and the value of f could remain at approximately 5.1–6.4 at any H2O2 dose (Fig. 6d). These results might be because the reaction rate between [FeII(C2O4)2]2− and H2O2 proceeded at a rate 3–4 orders of magnitude faster than that between FeII and H2O2. Therefore, the H2O2 addition could significantly enhance the removal of tartrazine with the UV/Fe0/Ox system.

Fig. 6
figure 6

a Tartrazine removal with the UV/Fe0/Ox system and a H2O2 addition. b Value of k for tartrazine removal with the UV/Fe0/Ox system and a H2O2 addition. c Tartrazine removal with the UV/Fe0 system and a H2O2 addition. d Value of k for tartrazine removal with the UV/Fe0 system and a H2O2 addition

Full size image

3.5 Comparing Various Catalysts and Chelating Agents

The performance of various catalysts adopted in the photolysis of ferric-oxalate complexes was compared. As shown in Fig. 7a, the tartrazine removal within 60 min was 35.9, 67.5, 69.4, 70.7, 81.1, and 86.4% with the UV/Fe2O3/Ox, UV/FeS/Ox, UV/Fe3O4/Ox, UV/FeIII/Ox, UV/Fe0/Ox, and UV/FeII/Ox systems, respectively. The increase in the value of k (103) on the tartrazine removal increased in the following order: Fe2O3 (10.7 min−1) < FeS (26.3 min−1) < Fe3O4 (42.4 min−1) < FeIII (47.3 min−1) < Fe0 (59.5 min−1) < FeII (77.1 min−1) (Fig. 7b). Compared with other heterogeneous iron catalysts, Fe0 could have a higher tartrazine removal and faster tartrazine removal rate. Compared with heterogeneous iron catalysts, Fe0 obtained a similar tartrazine removal efficiency and tartrazine removal rate. Therefore, Fe0 was a good heterogeneous iron catalyst for the photolysis of pollutants with the Ox system.

Fig. 7
figure 7

a Tartrazine removal with different Fe source systems. b Value of k for tartrazine removal with different Fe source systems. c Tartrazine removal with different chelating agent systems. d Value of k for tartrazine removal with different chelating agent systems. Reaction conditions: [catalysts] 0.2 g L−1, [chelating agent] 0.5 mM, pH 4, and [tartrazine] 20 mg L−1

Full size image

To clarify the photolysis performance of various chelating agents, the tartrazine removal with different UV/Fe0/chelating agents was compared. As depicted in Fig. 7c, the tartrazine removal within 60 min was 19.3, 58.7, 78.4, and 81.1% with the UV/Fe0/H3PO4, UV/Fe0/EDTA, UV/Fe0/NTA, and UV/Fe0/Ox systems, respectively. The increase in the value of k (103) on the tartrazine removal increased in the following order: H3PO4 < EDTA < NTA < Ox (59.5 min−1) (Fig. 7d). Thus, Ox was a good photolysis chelating agent.

3.6 Reusability of Fe0

Fe0 could be a good heterogeneous iron catalyst for the photolysis of tartrazine with the Ox system; thus, it is critical to examine the recyclability of the catalyst for long-term use. To assess the reusability of Fe0 in the UV/Fe0/Ox system, five consecutive experiments of tartrazine removal were conducted under the same reaction conditions. As presented in Fig. 8a, after five consecutive experiments, the tartrazine removal within 30 min was 86.9, 84.1, 84.2, 81.3, and 81.9%, respectively. The values of k (103) were 202.6, 187.2, 201.7, 165.2, and 176.4 min−1 in the 1st, 2nd, 3rd, 4th, and 5th runs, respectively (Fig. 8b). These results demonstrated that Fe0 showed good stability for long-term operation.

Fig. 8
figure 8

Repeated runs with the UV/pre-Fe0/Ox system: (a) tartrazine removal and (b) k values. Reaction conditions: [Fe0] 0.2 g L−1, [Ox] 0.5 mM, pH 4, and [tartrazine] 10 mg L−1

Full size image

4 Conclusion

In this study, Ox was adopted to enhance a UV/Fe0 system for the removal of tartrazine. Only 47.4% of tartrazine could be removed within 60 min with the UV/Fe0 system, while 72.4% of tartrazine could be removed within only 20 min with the UV/Fe0/Ox system. The effects of various factors, such as the Fe0 dose (0–0.8 g/L), Ox dose (0–2 mM), initial pH (2–6), and initial tartrazine concentration (2–30 mg/L), on the tartrazine removal were examined. The optimal Fe0 dose, Ox dose, and initial pH were 0.4 g/L, 1 mM, and pH 3, respectively. Quenching experiments demonstrated that •OH was the major radical for the removal of tartrazine, H2O2 was the major intermediate for •OH generation, and O2•− played a minor role in tartrazine removal. SEM-EDX, FTIR, and Mossbauer spectroscopy were conducted to explore the mechanism by which oxalate enhanced the performance of the UV/Fe0 system. Ox could inhibit the formation of iron (hydro)xides on the Fe0 surface, thus guaranteeing the reactivity of Fe0 during the reaction. Compared with Fe2O3, FeS, and Fe3O4, Fe0 was a good heterogeneous iron catalyst for the photolysis of pollutants with the Ox system. Compared with H3PO4, nitrilotriacetic acid, and ethylenediaminetetraacetic acid, Ox was also a good photolysis chelating agent. Finally, Fe0 showed good stability for long-term operation with the UV/Fe0/Ox system.