Introduction

In recent years, peracetic acid (PAA), an alternative disinfectant for chlorine in wastewater treatment, has attracted extensive attention and interest in advanced oxidation processes (AOPs)1. UV radiation2, heat3, ultrasound4, electrochemistry5, phosphate6, transition metals7,8,9,10,11 and carbonaceous materials12 have been confirmed to be able to activate PAA. Among these PAA activators, transition metals are considered as the most cost-effective and viable candidates due to their abundant natural abundance and efficiently catalytic performance13. Of which, Fe(II) and Co(II) are widely studied metals because of their most efficient activation to PAA7,8,13. Additionally, other transition metals (e.g., Cu(II), Mn(II), etc.) are also applied in PAA-based AOPs14,15, but the related researches are limited probably due to their low catalytic abilities. In our previous study, we found that HCO3−/CO32− or phosphate (PBS) could enhance PAA activation by Cu(II) because of the generation of Cu(II)-CO32− or Cu(II)-PBS complexes16,17,18. Very recently, Kim et al.19 reported that picolinic acid (PICA) could mediate Mn(II)-catalyzed PAA through the generation of Mn(II)-PICA complexes. Dong et al.20 also found that the degradation of atrazine (ATZ) was enhanced in Mn(II)/PAA process with the addition of ethylenediamine-N,N′-disuccinic acid (EDDS). These findings suggest that the introduction of complexing agent in Cu(II)/Mn(II)-mediated PAA systems can improve PAA activation.

Usually, hydroxyl radical (·OH) and organic radicals (RO·) are supposed to be the predominant reactive oxidants in transition metal-activated PAA systems9,21,22. However, some latest researches reported that except radical oxidation, non-radical oxidation processes are also responsible for the abatement of pollutants in these systems23,24. For instance, Kim et al.13 found that Fe(IV) could also be formed in Fe(II)/PAA process together with radical species, and RO· and Fe(IV) both contributed to the elimination of micropollutants in the initial kinetic phase. Liu et al.25 considered that Co(IV) rather than RO· was the dominant reactive oxidant for sulfonamides degradation in Co(II)-doped g-C3N4/PAA process. Additionally, in Mn(II)/EDDS/PAA system, Mn(V) has been verified to be the major reactive species for ATZ abatement20, because the introduction of ligand can enhance the stability of manganese intermediates. Hence, the contribution of high-valent metals in transition metal-catalyzed PAA systems cannot be ignored.

Mn, a common transition metal, has high natural abundance, rich oxidation states and environmental friendliness compared to Co and Cu. Compared with Fe, it can be used in a wide pH range without formation of sludge. Therefore, the activation of PAA by Mn(II) in the presence of nitrilotriacetic acid (NTA), a commonly used organic ligand, was investigated in this work. Previous researches have confirmed that NTA can coordinate with Mn(II) to produce a complex enhancing the treatment efficiency of Mn(II)/periodate (PI) and Mn(II)/peroxymonosulfate (PMS) processes26,27,28. Hence, the introduction of NTA in Mn(II)/PAA process may also be able to improve PAA activation, while the related publication is very limited. Sulfamethoxazole (SMX) was used as the target pollutant in this work, because it is a typical antibiotic that is detected frequently in groundwater, surface water and even tap water and cannot be completely eliminated by traditional wastewater treatment techniques22.

The main aims of this study were to: (1) evaluate the feasibility of NTA-improved treatment efficiency of Mn(II)/PAA process; (2) check the generation of high-valent manganese intermediates and distinguish the reactive oxidant for SMX abatement in Mn(II)/PAA/NTA process; (3) propose the possible degradation pathways of SMX in this process; (4) investigate the influence of reagent dosage, solution pH and water matrix on SMX elimination; (5) explore the applicability of Mn(II)/PAA/NTA system through degrading different kinds of contaminants.

Materials and methods

Materials

The chemicals used in this study were all analytical grade and are shown in Text S1 in supporting information.

Batch experiments

All batch experiments were performed in a 250 mL glass reactor at room temperature (25 °C) equipped with constant magnetic stirring. Predetermined concentrations of PAA, NTA and target contaminants were first added into the glass reactor. Subsequently, the solution pH was immediately adjusted by 100 mM H2SO4 solution and 100 mM NaOH solution. Finally, the experiments were initiated as soon as the manganese acetate solution was added. At designated time intervals, 1 mL reaction solution was taken out and immediately mixed with excess ascorbic acid to quench the residual oxidants. All batch experiments were conducted at least twice to minimize errors.

Analysis

The analytic methods of SMX, methyl phenyl sulfone (PMSO2), methyl phenyl sulfide (PMSO) and other contaminants are presented in Text S2. The detected information about SMX degradation products and residual PAA concentration in reaction system can be found in our published papers29,30. The UV–vis absorption spectrum of Mn(III) species was monitored by an UV–vis spectrophotometer (UV4802, UNICO) at 200–600 nm. The solution pH was measured using a pH meter (PHS-3C, Leici).

Results and discussion

SMX degradation in Mn(II)/PAA/NTA process

Figure 1 presents the abatement of SMX in different systems. In PAA alone and Mn(II)/PAA processes, negligible SMX removal within 1 min was both observed, suggesting that Mn(II) could not efficiently activate PAA to produce reactive oxidants. However, when NTA was added into Mn(II)/PAA process, a rapid SMX degradation was obtained and 79.4% SMX was removed within only 5 s. Meanwhile, the degradation of SMX in several control groups including NTA alone, Mn(II)/NTA and PAA/NTA was conducted, as shown in Fig. 1. In these three systems, SMX was hardly degraded, indicating that both NTA and Mn(II)-NTA complexes had no reactivity toward SMX as well as NTA was unable to activate PAA. Through comparing the above results, we could conclude that the addition of complexing ligand NTA in Mn(II)/PAA process could actually improve PAA activation by the formed Mn(II)-NTA complex. As a result, the rapid elimination of SMX in Mn(II)/PAA/NTA process was probably ascribed to the formed reactive oxidant by the activated PAA. Since the used PAA in this work contains H2O2, SMX abatement in Mn(II)/H2O2/NTA process was also studied to clarify the possible contribution of H2O2 in Mn(II)/PAA/NTA process. As depicted in Fig. 1, no SMX elimination was observed in Mn(II)/H2O2/NTA process, which suggested that H2O2 was an inefficient oxidant in Mn(II)/PAA/NTA process for SMX removal.

Fig. 1
figure 1

Degradation of SMX in different systems. Experimental conditions: [PAA]0 = 0.2 mM, [H2O2]0 = 0.48 mM, [NTA]0 = 25 μM, [Mn(II)]0 = 5 μM, [SMX]0 = 10 μM, pH0 = 6.5, T = 25 °C.

Full size image

Identification of reactive species

Identification of active radicals

Since ·OH and organic radicals (i.e., CH3C(O)O· and CH3C(O)OO·) have been reported to be the major active radicals for contaminants abatement in PAA-based AOPs systems1,21,22. Tert-butanol (TBA) and methanol (MeOH) were introduced into Mn(II)/PAA/NTA process to identify the possible contribution of aforementioned radicals, because TBA could selectively quench ·OH (kTBA/·OH = 6 × 108 M−1 s−131) while MeOH could efficiently scavenge ·OH and organic radicals. Although the rate constants of MeOH with organic radicals are unknown, the efficient scavenging of organic radicals by MeOH had already been reported8. As shown in Fig. 2a, TBA had an insignificant impact on SMX elimination in Mn(II)/PAA/NTA process, however, MeOH induced a slight inhibition on SMX removal, indicating that organic radicals rather than ·OH might participate in the abatement of SMX in Mn(II)/PAA/NTA process.

Fig. 2
figure 2

Effect of MeOH and TBA on SMX degradation in Mn(II)/PAA/NTA process (a); degradation of SMX by Mn(III)-NTA alone and Mn(III)-NTA/PAA within 60 s (b); effect of EtOH on SMX degradation in Mn(II)/PAA/NTA process (c); degradation of PMSO and formation of PMSO2 in Mn(II)/PAA/NTA process (d). Experimental conditions: [PAA]0 = 0.2 mM, [NTA]0 = 25 μM, [Mn(II)]0 = [Mn(III)-NTA]0 = 5 μM, [MeOH]0 = [TBA]0 = [EtOH]0 = 100 mM, [PMSO]0 = 10 μM, [SMX]0 = 10 μM, pH0 = 6.5, T = 25 °C.

Full size image

Identification of Mn(III) species

Usually, once Mn(III) is formed, it will undergo disproportionation via Eq. (1), producing Mn(II) and less reactive MnO232. However, Mn(III) can be stabilized with the presence of complexing agents through forming Mn(III)-complexes with ligands33,34, which is considered as a potential oxidant for the rapid elimination of contaminants in previous studies35,36,37. Hence, in this work, the presence of NTA might improve the generation of stabilized Mn(III) species (i.e., Mn(III)-NTA complex) in Mn(II)/PAA/NTA process26,33. In order to verify the formation of Mn(III) species in this process, UV–vis absorption spectrum of reaction solution at 200–600 nm was detected, and the results are shown in Fig. S1. Different from previous study26, the characteristic peak of Mn(III) species at ~ 270 nm was extremely weak in Mn(II)/PAA/NTA process and rapidly disappeared after 5 s. This result indicated that the stabilized Mn(III) species might generate in Mn(II)/PAA/NTA process but transform to other Mn species rapidly. To investigate the possible contribution of Mn(III)-NTA complex to SMX abatement in Mn(II)/PAA/NTA process, Mn(III)-NTA complex was ex-situ prepared using Mn(II)-NTA complex and permanganate to degrade SMX26. As depicted in Fig. 2b, Mn(III)-NTA complex alone could hardly degrade SMX, hence, the formed Mn(III)-NTA complex in Mn(II)/PAA/NTA process was not the main reactive oxidant for SMX abatement. However, the combination of PAA and Mn(III)-NTA complex induced 98.4% SMX elimination in 60 s. These results suggested that in Mn(II)/PAA/NTA process, the generated Mn(III)-NTA complex could not directly oxidize SMX while it could react with PAA producing highly reactive species for SMX abatement. The residual PAA concentration in Mn(II)/PAA/NTA and Mn(III)-NTA/PAA processes were thus monitored, respectively. As shown in Fig. S2, after reaction for 60 s, approximately 30% PAA was consumed in these two systems, which further confirmed the above discussion.

$${\text{2Mn(III) + 2H}}_{{2}} {\text{O }} \to {\text{ Mn(II) + MnO}}_{{2}} {\text{ + 4H}}^{ + }$$
(1)

Identification of high-valent Mn species

Since the formed Mn(III)-NTA would further react with PAA in Mn(II)/PAA/NTA process, it might transform to a higher valence state, and other high-valent Mn species (e.g., Mn(IV), Mn(V) and Mn(VI)) were regarded as the potential reactive Mn intermediates for SMX abatement in Mn(II)/PAA/NTA process26,38. Hence, the possible contribution of these high-valent Mn species to SMX removal were investigated. Since Mn(IV) usually existed in the form of less reactive MnO2, the abatement of SMX in MnO2/PAA and MnO2/PAA/NTA processes were also investigated to distinguish the possible contribution of Mn(IV) in Mn(II)/PAA/NTA process. As depicted in Fig. S3, less than 10% SMX elimination was observed in the above two systems, suggesting that Mn(IV) had limited contribution to SMX degradation in Mn(II)/PAA/NTA process. In addition, Mn(VI) only existed under alkaline condition39. Therefore, Mn(V) was supposed to be the only reactive Mn intermediate for SMX abatement in Mn(II)/PAA/NTA process. Ethanol (EtOH) is an efficient scavenger for ·OH (kEtOH/·OH = 2.2 × 108 M−1 s−140), while recent works reported that EtOH could efficiently scavenge high-valent metal species41,42. As "Identification of active radicals" section had proved that ·OH played an insignificant role in SMX elimination in Mn(II)/PAA/NTA process, hence, the introduction of EtOH could probe if Mn(V) contributed to SMX abatement in this process. As presented in Fig. 2c, the addition of EtOH in Mn(II)/PAA/NTA process induced a strong inhibition on SMX abatement, which was much more severe than that of MeOH (Fig. 2a). This result suggested that Mn(V) was the dominate reactive oxidant for SMX elimination in Mn(II)/PAA/NTA process. In order to further identify the contribution of Mn(V) in Mn(II)/PAA/NTA process, PMSO, a chemical probe to distinguish the contribution of Mn(V) and reactive radicals, was introduced into this system. Because PMSO would be oxidized to PMSO2 in the presence of Mn(V), whereas reactive radicals would decompose PMSO to hydroxylated and/or polymeric products26,38,43. Figure 2d shows the PMSO degradation and PMSO2 generation as well as the calculated yield rate of PMSO2 (η, the ratio of yielded PMSO2 concentration to degraded PMSO concentration). The concentration of the formed PMSO2 increased with the decrease of PMSO concentration and η kept at ~ 90%. The above findings indicated that both reactive radicals and Mn(V) participated in the decomposition of PMSO, and Mn(V) was the dominate working oxidant in Mn(II)/PAA/NTA process, which further supported the results of MeOH and EtOH scavenging experiments. Similar conclusion was also obtained in Mn(II)/PI/NTA process where Mn(V) was the predominate working oxidant in this process27,44.

Reaction mechanism in Mn(II)/PAA/NTA process

Based on the above discussion, the possible reaction mechanism in Mn(II)/PAA/NTA process was put forward in Fig. 3. Firstly, Mn(II) and NTA coordinated to produce Mn(II)-NTA complex (i.e., MnNTA− shown in Fig. S4), as described in Eq. (2)27,38. Subsequently, the electron transfer occurred between Mn(II)-NTA complex and PAA, generating CH3C(O)O· and Mn(III)-NTA complex (Eq. (3)). The generated Mn(III)-NTA complex would further react with PAA to produce Mn(V) through two-electron transfer (Eq. (4)) and/or CH3C(O)OO· via one-electron transfer (Eq. (5))27,44. These generated reactive species including organic radicals (CH3C(O)O· and CH3C(O)OO·) and Mn(V) contributed to SMX abatement in Mn(II)/PAA/NTA process, and Mn(V) played a major role in SMX elimination.

$${\text{Mn(II) + NTA }} \to {\text{ Mn(II)}} - {\text{NTA}}$$
(2)
$${\text{Mn(II)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)OOH }} \to {\text{ Mn(III)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)O}}^{ \cdot } {\text{ + OH}}^{ - }$$
(3)
$${\text{Mn(III)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)OOH }} \to {\text{ Mn(V)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)O}}^{ - } {\text{ + OH}}^{ - }$$
(4)
$${\text{Mn(III)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)OOH }} \to {\text{ Mn(II)}} - {\text{NTA + CH}}_{{\text{3}}} {\text{C(O)OO}}^{ \cdot } {\text{ + H}}^{ + }$$
(5)
Fig. 3
figure 3

Possible reaction mechanism in Mn(II)/PAA/NTA process.

Full size image

Possible degradation pathways of SMX

Six degradation products of SMX were detected in Mn(II)/PAA/NTA process. As a result, four probable SMX degradation pathways including hydroxylation, amino oxidation, coupling reaction and bond cleavage were proposed, as presented in Fig. 4. Hydroxylation (pathway (1)), a common SMX transformation pathway in AOPs8,22, might occur on the aniline moiety of SMX, generating product m/z 270 (hydroxyl-SMX). The amino moiety on the aromatic ring of SMX might be oxidized to form m/z 284 (nitro-SMX) (pathway (2)) due to its electron-rich nature. Besides, the S–N bond in SMX suffered a cleavage (pathway (3)) in Mn(II)/PAA/NTA process, resulting in the formation of products m/z 99 (3-amino-5-methylisoxazole) and m/z 158 (4-amino benzene sulfonic acid) that was not identified in this work22. Coupling reaction (pathway (4)) might also occur during SMX decomposition in Mn(II)/PAA/NTA process based on the identification of three dimeric products of SMX. Product m/z 519 (hydroxylated azo-SMX) might be generated via pathway (4) between N-centered radicals formed from hydroxyl-SMX and SMX, respectively, while m/z 503 (azo-SMX) might be produced through this pathway between two N-centered radicals formed from SMX8.

Fig. 4
figure 4

Possible degradation pathways of SMX in Mn(II)/PAA/NTA process: (1) hydroxylation; (2) nitration; (3) bond cleavage; (4) coupling reaction. The product in “[ ]” was not detected in this study.

Full size image

Influencing factors

Effect of initial pH

Solution pH could affect the existing form of SMX and PAA in Mn(II)/PAA/NTA process, which could further influence SMX removal. Therefore, the effect of initial pH on the abatement of SMX in this process was explored. Based on the pKa of SMX (pKa1 = 1.8 and pKa2 = 5.6)45, neutral form and deprotonated form of SMX are coexisted at pH 5.0, while with the increase of pH from 6.5 to 11.0, deprotonated form is the primary existing form of SMX in Mn(II)/PAA/NTA process. In this studied pH range, the existing form of SMX might have no impact on its removal. As depicted in Fig. 5a, acidic condition was more beneficial to SMX removal than alkaline condition in this system, which was probably due to that the abundant H+ under acidic condition would promote the generation of reactive species via Eqs. (3) and (4). However, in alkaline condition, the degradation of SMX was inhibited severely, because PAA existed mainly in the form of PAA− at pH 9.5 and 11.0 due to its pKa value of 8.2 probably affecting the formation of reactive species19,46. Besides, abundant OH− under alkaline condition would restrain the formation of reactive oxidant in Mn(II)/PAA/NTA process through Eqs. (3) and (4). Since the initial pH of solution was adjusted by H2SO4 and NaOH in this work, the solution pH might change after reaction. Hence, the solution pH after reaction in this process was determined. As shown in Table S2, the solution pH after reaction at different initial pH values was almost unchanged, probably because of the short reaction time (i.e., 1 min). Consequently, the pH change of solution had no impact on SMX removal in Mn(II)/PAA/NTA process.

Fig. 5
figure 5

Effect of initial pH (a), PAA dosage (b), Mn(II) dosage (c) and NTA dosage (d) on SMX degradation in Mn(II)/PAA/NTA process. Experimental conditions: [PAA]0 = 0.2 mM except (b), [Mn(II)]0 = 5 μM except (c), [NTA]0 = 25 μM except (d), pH0 = 6.5 except (a), [SMX]0 = 10 μM, T = 25 °C.

Full size image

Effect of reagent concentration

The dosage of reagents could significantly affect the treatment efficiency of Mn(II)/PAA/NTA process, different concentrations of PAA, Mn(II) and NTA were thus utilized to assess their effects on SMX elimination in this process, respectively. As presented in Fig. 5b and c, the abatement of SMX enhanced as the concentrations of PAA and Mn(II) increased, because more PAA and Mn(II) could generate more reactive oxidant for SMX abatement in Mn(II)/PAA/NTA process20. With the increase of NTA dosage, more reactive Mn species could be stabilized through complexation, leading to the enhancement in SMX removal, as depicted in Fig. 5d. Taking removal efficiency and operational cost into consideration, 0.2 mM of PAA, 5 μM of Mn(II) and 25 μM of NTA were recommended in Mn(II)/PAA/NTA system, respectively, because very limited increase in SMX degradation occurred when their concentrations increased further.

Effect of water matrix

The effect of common water matrix (e.g., Cl−, SO42−, NO3−, HCO3−, fulvic acid (FA), Ca2+ and Mg2+) on SMX abatement by Mn(II)/PAA/NTA process was investigated. As depicted in Fig. 6a and b, SO42− and NO3− exhibited negligible effect on SMX abatement in Mn(II)/PAA/NTA process, while Cl− and HCO3− inhibited SMX degradation (Fig. 6c and d), and this inhibition effect was more obvious as their concentrations increased. The suppression effect of Cl− in Mn(II)/PAA/NTA process was probably attributed to its competition with SMX for organic radicals (Eq. (6))47. Since HCO3− could coordinate with Mn(II) producing Mn(II)-HCO3− complex28, it would inhibit the complexation of Mn(II) with NTA, which reduced the formation of Mn(II)-NTA complex and subsequent Mn(V). The presence of FA in Mn(II)/PAA/NTA process suppressed SMX removal, and this inhibition effect enhanced with the elevating FA concentration, as depicted in Fig. 6e. Possible explanation for this finding was that FA might not only affect the formation of Mn(II)-NTA complex but also compete with SMX for reactive oxidant44,48. As shown in Fig. 6f, the presence of Ca2+ significantly suppressed the elimination of SMX in Mn(II)/PAA/NTA process, but a slight inhibition on SMX abatement was observed in this process with the addition of Mg2+ (Fig. 6g). Due to the stronger complexing ability of Ca2+ than that of Mg2+ toward NTA, they might compete with Mn(II) for NTA reducing the formation of Mn(II)-NTA complex and subsequent high-valent Mn species, resulting in a stronger inhibition on SMX removal by Ca2+ than that by Mg2+28,49. Finally, the SMX removal by Mn(II)/PAA/NTA process in real waters was conducted. The two real waters were collected from Tuojiang River near our university and Xi Lake in our campus, respectively, and their water parameters are shown in Table S3. As depicted in Fig. 6h, SMX elimination in these two real waters were both inhibited compared with that in ultrapure water, which was probably due to the interference of water matrix.

$${\text{CH}}_3{\text{C(O)OO}}^{ \cdot } {\text{ + Cl}}^{ - } {\text{ + H}}^{ + } { } \to {\text{ CH}_3{\text{C(O)OOH + Cl}}}^{ \cdot }$$
(6)
Fig. 6
figure 6figure 6

Effect of SO42− (a), NO3− (b), Cl− (c), HCO3− (d), FA (e), Ca2+ (f) and Mg2+ (g) on SMX degradation in Mn(II)/PAA/NTA process; SMX degradation by Mn(II)/PAA/NTA process in real waters (h). Experimental conditions: [PAA]0 = 0.2 mM, [Mn(II)]0 = 5 μM, [NTA]0 = 25 μM, pH0 = 6.5, [SMX]0 = 10 μM, T = 25 °C.

Full size image

Degradation of other organics in Mn(II)/PAA/NTA process

To further assess the application prospect of Mn(II)/PAA/NTA process, five different kinds of organics were selected to investigate their degradation efficiency in this process. As depicted in Fig. 7, the removal efficiency of diclofenac (DCF), aniline (AN), rhodamine B (RhB), methyl orange (MO) and orange G (OG) in Mn(II)/PAA/NTA process within 60 s were 97.1%, 34%, 3.1%, 2.2% and 2.1%, respectively, suggesting that this process possessed high selectivity toward different organics. Among these contaminants, SMX, DCF and AN all contain electron-rich group (i.e., aniline group), and Mn(V) was the predominant working oxidant in this process. These results suggested that Mn(V) had a higher reactivity toward organics with electron-rich groups50.

Fig. 7
figure 7

Degradation of other contaminants in Mn(II)/PAA/NTA process. Experimental conditions: [PAA]0 = 0.2 mM, [Mn(II)]0 = 5 μM, [NTA]0 = 25 μM, [DCF]0 = [AN]0 = 10 μM, [RhB]0 = [OG]0 = [MO]0 = 20 mM, pH0 = 6.5, T = 25 °C.

Full size image

Conclusion

In this work, NTA was found to be able to improve SMX abatement by Mn(II)/PAA process through complexing with Mn(II) to produce Mn(II)-NTA complex which could accelerate PAA activation. In Mn(II)/PAA/NTA process, organic radicals (i.e., CH3C(O)O· and CH3C(O)OO·) and Mn(V) both contributed to SMX elimination and Mn(V) was verified to be the predominant reactive oxidant. Hydroxylation, nitration, coupling reaction and bond cleavage were the four probable degradation pathways of SMX in this process according to the six identified degradation products. Mn(II)/PAA/NTA system exhibited high treatment efficiency in acidic and neutral conditions, and increasing reagent concentration was also beneficial to SMX removal in this process. The presence of SO42− and NO3− in Mn(II)/PAA/NTA process had little impact on SMX abatement, while Cl−, HCO3−, FA, Ca2+ or Mg2+ could inhibit SMX removal and the inhibition effect increased with their increasing dosage. Mn(II)/PAA/NTA process showed a high reactivity toward organics with electron-rich groups. Although PAA, NTA and acetate are remained in the reaction solution, they can be degraded by microorganism, which may not cause the secondary pollution if this process is used as a pretreatment of wastewater followed by biological treatment.