UV absorbance and electron donating capacity as surrogate parameters to indicate the abatement of micropollutants during the oxidation of Fe(II)/PMS and Mn(II)/NTA/PMS

In this study, the relative residual UV absorbance (UV254) and/or electron donating capacity (EDC) was investigated as a surrogate parameter to evaluate the abatement of micropollutants during the Fe(II)/PMS and Mn(II)/NTA/PMS processes. In the Fe(II)/PMS process, due to the generation of SO4•- and •OH at acidic pH, UV254 and EDC abatement was greater at pH 5. In the Mn(II)/NTA/PMS process, UV254 abatement was greater at pH 7 and 9, while EDC abatement was greater at pH 5 and 7. This was attributed to the fact that MnO2 was formed at alkaline pH to remove UV254 by coagulation, and manganese intermediates (Mn(V)) were formed at acidic pH to remove EDC via electron transfer. Due to the strong oxidation capacity of SO4•-, •OH and Mn(V), the abatement of micropollutants increased with increasing dosages of oxidant in different waters in both processes. In the Fe(II)/PMS and Mn(II)/NTA/PMS processes, except for nitrobenzene (∼23% and 40%, respectively), the removal of other micropollutants was greater than 70% when the oxidant dosages were greater in different waters. The linear relationship between the relative residual UV254, EDC and the removal of micropollutants was established in different waters, showing a one-phase or two-phase linear relationship. The differences of the slopes for one-phase linear correlation in the Fe(II)/PMS process (micropollutant-UV254: 0.36-2.89, micropollutant-EDC: 0.26-1.75) were less than that in the Mn(II)/NTA/PMS process (micropollutant-UV254: 0.40-13.16, micropollutant-EDC: 0.51-8.39). Overall, these results suggest that the relative residual UV254 and EDC could truly reflect the removal of micropollutants during the Fe(II)/PMS and Mn(II)/NTA/PMS processes.


Introduction
In recent years, micropollutants such as pharmaceuticals and personal care products, endocrine disruptors, and pesticides have been found in drinking water and secondary wastewater e uent (Song et al., 2017). These micropollutants are primarily composed of phenolic and amine compounds, which have three carcinogenic effects (i.e., mutagenic, teratogenic, and carcinogenic) (Ternes, 1998) An advanced oxidation process such as persulfate activated oxidation is an effective approach for reducing micropollutants (Rastogi et al., 2009;Ji et al., 2015). The peroxymonosulfate (PMS) structure is asymmetric, with only one SO 3 group attached to its peroxy group, which has greater reactivity than peroxydisulfate (PDS) (Gao et al., 2021). Recent  In addition, a current study found that manganese intermediates with high oxidation capacity were formed by Mn(II)/complexing agent/persulfate that removed organic contaminants signi cantly (Gao et al., 2021;Hu et al., 2022). Therefore, activating PMS to generate reactive intermediate species for the abatement of micropollutants is a feasible method. However, when adding oxidants to remove micropollutants in actual water, the water matrix (i.e., natural organic matter) competed with the micropollutants for the oxidant (Song et al., 2017). Therefore, the abatement of micropollutants may be related to the degradation of certain reactive groups of natural organic matter.
Currently, a number of parameters such as UV absorbance at 254 nm (UV 254 ), total uorescence (TF), and electron donating capacity (EDC) re ect the reactive groups of natural organic matter (Shang et

Chemicals and reagents
All chemicals were analytical grade or higher quality and obtained from commercial suppliers (see Table   S1 for details). The selected micropollutants comprised of phenols, sulfonamides and heterocyclic organic compounds, including phenol, bisphenol A (BPA), sulfamethazine (SMZ), sulfamethoxazole (SMX), atrazine (ATZ) and nitrobenzene (NB) (see Table S2 for details). All solutions were prepared with ultrapure water (18 MΩcm) from the water puri cation system (Milli-Q Biocel, Millipore, USA). In the Fe(II)/PMS and Mn(II)/NTA/PMS processes, the buffer solutions at pH 5 comprised of 10 mM sodium acetate, and the buffer solutions at pH 7 and 9 comprised of 10 mM sodium tetraborate. The ozone generated by a ozone generator (CMG3-5, Innovatech, Germany) was dissolved into ultrapure water at 4 ℃ for 1 h to prepare ozone stock solution, and the ozone stock solution was used immediately.

Water samples
Standard humic acid (HA) purchased from Aladdin Reagent Company was used as the natural organic component of synthetic water. The synthetic water was a 4.25 mgC/L HA solution. The details of groundwater and surface water are shown in Table S3. After the groundwater and surface water were retrieved, they were ltered with 0.45 µm microporous membranes. Then they were stored at 4 ℃ and used within a week.

Experiment process
This experiment was performed at room temperature (25 ± 1 ℃) at pH 5, 7, and 9. The humic acid solutions for synthetic water and the authentic water samples (i.e., groundwater and surface water) spiked with 2 µM micropollutants (i.e., SMZ, ATZ, SMX, phenol, BPA, and NB) were dosed with different oxidant dosages at different pH values (10 mM sodium acetate buffer at pH 5 or sodium tetraborate buffer at pH 7 and 9) and the original pH (authentic water samples only). In the Fe(II)/PMS process,

Analytical methods
The concentration of selected micropollutants was determined by high performance liquid chromatography (HPLC) (Waters 1525, Waters, USA). The detailed selected micropollutant determination methods are shown in Table S4. The concentration of aqueous ozone stock solution was measured by spectrophotometer (Cary 3000, Varian, USA) with absorbance at 260 nm (ε = 3200 M − 1 cm − 1 ) (Elovitz and von Gunten, 1999). The PMS stock solution was prepared and determined as described in Text S1. The EDC was determined using the post-column derivatization radical oxidation method (Chon et   For groundwater and surface water (Figs. 1c and 1d), with the increase of Fe(II) and PMS dosages (0-100 µM), UV 254 and EDC were abated by 99% and 50% in groundwater, and 48% and 51% in surface water, respectively. The abatement of UV 254 in groundwater was signi cantly greater than that in surface water, which was due to the presence of greater amounts of Fe(II) and Mn(II) in groundwater (Table. S3).
Therefore, more Fe ( Moreover, the abatement of UV 254 at pH 9 was greater than that at pH 5 in synthetic water (i.e., 63% at pH 9, 41% at pH 5) with a PMS dosage of 80 µM (Fig. 2a). This was attributed to the fact that larger amounts of MnO 2 were generated at weak alkaline pH and effectively removed UV 254 by coagulation (Huangfu et al., 2013;Yu et al., 2022). However, when the PMS dosage was 80 µM, the abatement of EDC at pH 5 was greater than that at pH 9 in synthetic water (i.e., 78% at pH 5, 61% at pH 9) (   in ozone without t-BuOH, 38% in the Fe(II)/PMS process, 61% in the Mn(II)/NTA/PMS process) (Fig. 3b).
This result was due to the fact that NOM moieties with EDC were oxidized by SO 4 •− and •OH via one- In the Mn(II)/NTA/PMS process (Fig. 4b), in phase I, the slopes for the UV 254 -EDC correlations were also close to 1 in synthetic and surface water (i.e., 0.96 and 0.90, respectively). This was attributed to the oxidation of activated organics with EDC and absorbance at 254 nm (e.g., activated aromatic moieties) by manganese intermediates in phase I (Tian et al., 2022). However, the degradation of EDC was signi cantly greater than that of UV 254 in phase II, with slopes for UV 254 -EDC correlations of 0.04 and 0.16 in synthetic and surface water, respectively. This nding was due to the fact that the manganese intermediates oxidized phenolic moieties to form quinone-type moieties in phase II, which absorbed at 254 nm (Rouge et al., 2020). Due to the presence of the water matrix in surface water, the removal of micropollutants was relatively less in the Fe(II)/PMS process. However, for surface water treated by the Mn(II)/NTA/PMS process, the removal of BPA and phenol was not signi cantly affected by NOM, however, the removal of other micropollutants was relatively less. Figure S1 shows the abatement of selected micropollutants (e.g., SMZ and ATZ) treated by the During the Fe(II)/PMS process (Figs. 6a, 6b and 6c), in phase I, the slopes for the micropollutant-UV 254 correlations of SMZ, ATZ and SMX were greater than 1 in synthetic water, groundwater and surface water, indicating that SMZ, ATZ and SMX kinetically outcompeted the oxidation of NOM moieties absorbing at 254 nm by •OH and SO 4 •− . The slopes for the micropollutant-UV 254 correlations of phenol (i.e., 1.53 in synthetic water, 1.01 in groundwater, and 0.90 in surface water) in three different waters were close to 1 in phase I, indicating that the abatement of phenol increased proportionally with the decrease of relative residual UV 254 . However, the abatement of BPA and NB was less than the decrease of relative residual UV 254 in phase I in different waters, indicating that the effects of •OH and SO 4 •− on UV 254 were greater than BPA and NB. The differences between the correlations for these micropollutants were relatively small but in agreement with the relative order of their •OH and SO 4 •− reactivity (i.e., SMZ > ATZ > SMX > phenol > BPA > NB). In phase II, when UV 254 degraded completely, the abatement of micropollutants was less than 100% in synthetic water and groundwater, indicating that NOM moieties absorbing at 254 nm had the advantage in kinetically competing with micropollutants for •OH and SO 4 •− . However, when the relative residual UV 254 was 0.5, the other micropollutants (SMZ, ATZ, SMX, phenol and BPA) removals exceeded 70% in surface water, which was different from the nding in synthetic water and groundwater.

Removal of micropollutants during oxidation in the Fe(II)/PMS and Mn(II)/NTA/PMS processes
This was attributed to the different components in natural organic matter in surface water, compared with synthetic water and groundwater.
In the Mn(II)/NTA/PMS process (Figs. 6d and 6e), the removal of BPA and phenol showed a one-phase linear relationship with the relative residual UV 254 in synthetic and surface water. The removal of BPA and phenol was signi cantly greater than that of UV 254 , indicating that BPA and phenol had the advantage in kinetically competing with NOM for manganese intermediates in synthetic and surface water. The removal of SMZ, ATZ, SMX and NB showed a two-phase linear relationship with the relative residual UV 254 in synthetic and surface water. In phase I, the slopes for the micropollutant-UV 254 correlations of SMZ were close to 1 in synthetic (1.17) and surface water (0.99), indicating that the abatement of SMZ increased proportionally with the decrease of relative residual UV 254 in phase I. The slopes for the micropollutant-UV 254 correlations of ATZ, SMX and NB were less than 1 in synthetic and surface water in phase I, indicating that NOM moieties absorbing at 254 nm kinetically outcompeted the oxidation of ATZ, SMX and NB for manganese intermediates. The differences between the correlations of these four micropollutants (SMZ, ATZ, SMX and NB) were relatively small, but in agreement with the relative order of their removal e ciencies by manganese intermediates (i.e., SMZ > ATZ > SMX > NB). In phase II, when the activated organic moieties were consumed, the residual non-activated organic moieties were abated slowly by manganese intermediates, resulting in the rapid degradation of these four micropollutants in synthetic and surface water.  Tables S7 and S8. For synthetic water and groundwater treated by the Fe(II)/PMS process ( Fig. 7a and 7b), the removal of micropollutants showed a one-phase linear correlation with the relative residual EDC. In synthetic water (Fig. 7a) For synthetic and surface water treated by the Mn(II)/NTA/PMS process (Figs. 7d and 7e), the differences between the correlations for micropollutants were relatively large, which agrees with the order of their removal e ciencies by manganese intermediates (i.e., BPA > phenol > SMZ > ATZ > SMX > NB). The removal of BPA and phenol showed a one-phase linear relationship with EDC degradation in synthetic and surface water. The removal of BPA and phenol was signi cantly greater than EDC degradation, indicating that BPA and phenol kinetically outcompeted the oxidation of NOM moieties with EDC. The removal of SMZ, ATZ, SMX, and NB also showed a one-phase or two-phase linear relationship with the degradation of EDC in synthetic and surface water. In phase I, SMZ, ATZ, SMX, and NB degraded slowly relative to the abatement of EDC, indicating that NOM moieties with EDC kinetically outcompeted the oxidation of micropollutants for manganese intermediates. When EDC was consumed in the initial phase, these micropollutants became competitive, with a higher slope relative to the relative residual EDC in phase II.

The linear relationship between the relative residual EDC and the removal of micropollutants
In summary, the relative residual UV 254 and EDC in different waters displayed linear correlation with the abatement of selected micropollutants in the Fe(II)/PMS and Mn(II)/NTA/PMS processes. Therefore, the relative residual UV 254 and EDC could be used as surrogate parameters for the abatement of micropollutants during the oxidation of different waters in the Fe(II)/PMS and Mn(II)/NTA/PMS processes.

Conclusions
During the oxidation of synthetic water, groundwater and surface water in the Fe(II)/PMS and Mn(II)/NTA/PMS processes, the relative residual UV 254 and EDC showed a linear relationship with the selected micropollutants. This result is summarized as follows: The abatement of UV 254 and EDC at pH 5 and 7 was greater than that at pH 9 in the Fe(II)/PMS process. In the Mn(II)/NTA/PMS process, the abatement of UV 254 at pH 5 and 7 was greater than that at pH 9, while the abatement of EDC at pH 7 and 9 was greater than that at pH 5.
For synthetic water, groundwater and surface water treated by the Fe(II)/PMS and Mn(II)/NTA/PMS processes, with an increase in oxidant dosages, the removal of selected micropollutants increased gradually. The abatement of micropollutants at pH 5 and 7 was greater than that at pH 9 in the