Peroxymonosulfate Activation by Co-Doped Magnetic Mn3O4 for Degradation of Oxytetracycline in Water

12 Co-doped magnetic Mn3O4 was synthesized by the solvothermal method and 13 adopted as an effective catalyst for the degradation of oxytetracycline (OTC) in water. 14 Synergistic interactions between Co-Mn3O4 and Fe3O4 not only resulted in the 15 enhanced catalytic activity through the activation of peroxymonosulfate (PMS) to 16 degrade OTC, but also made Fe3O4/Co-Mn3O4 easy to be separated and recovered 17 from aqueous solution. 94.2% of OTC could be degraded within 60 min at an initial 18 OTC concentration of 10 mg/L, catalyst dosage of 0.2 g/L and PMS concentration of 19 10 mM, the high efficiency of which was achieved in a wider pH range of 3.0-10.0. 20 The free radical quenching experiments showed that O2 •radicals and 1 O2 played the 21 main role in the degradation process. Co 3+ , Co 2+ , Fe 2+ , Fe 3+ , Mn 4+ , Mn 3+ and Mn 2+ on 22 Fe3O4/Co-Mn3O4 were identified as catalytic sites based on XPS analysis. Eventually, 23 the intermediates of OTC degradation were examined and the possible decomposition 24 pathways were proposed. The excellent catalytic performance of Fe3O4/Co-Mn3O4 not 25


Introduction
With people's increasing attention to environmental health, pharmaceutical and 9 personal care products (PPCPs) are widely used in medicines to prevent and treat 10 human and animal diseases and as personal care products in daily life [1,2], resulting 11 in the discharge of a large amount of wastewater containing PPCPs and their 12 metabolites into the environment. PPCPs have been considered as one of the 13 emerging pollutants and one of the biggest culprits affecting the water environment 14 [3][4][5]. Therefore, there is increasing concern about the harmful impacts of PPCPs and 15 their metabolites on microorganisms [6,7], plants [8], animals [9] and human beings technologies for the removal of OTC in water. At present, there have been many 1 reports on the methods of removing OTC from water bodies. At present, the methods 2 to remove oxytetracycline from water mainly include physical adsorption [16], Fenton 3 -like oxidation [17], electrochemical catalytic oxidation [18], photocatalysis [19], etc. 4 Sulfate radical-based advanced oxidation processes (SR-AOPs) are widely used 5 for oxidizing and degrading pollutants [20]. SR-AOPs can produce highly reactive 6 species that react directly with pollutants to oxidize them into less toxic compounds 7 even harmless small molecule compounds such as CO 2 and H 2 O. The SO 4 •and • OH 8 radicals produced by SR-AOPs exhibit excellent oxidation capacity in wastewater 9 treatment owing to their higher oxidation potential (2.5-3.1 V and 1.8-2.7 V vs. NHE, 10 respectively) [21,22], high selectivity and wide pH response range. In general, SO 4 •-

11
radical is produced via peroxymonosulfate (PMS) activation with various activation 12 methods, such as heat [23], UV [24,25], electro-activation [26], transition metal [27, 13 28] and ultrasound [29]. The transition metal activation is usually used because of its 14 low equipment requirements, low cost and more flexibility. 15 Zero-valent transition metals and their ions (Co 2+ , Cu 2+ , Mn 2+ , Ni 2+ , Fe 2+ , Fe 3+ , 16 etc.) are often used for the activation of PMS to oxidize pollutants. Among them, 17 Mn 3 O 4 is concerned because they contain multiple valence states [30]. The rapid 18 redox transformation between Mn 2+ , Mn 3+ and Mn 4+ plays a key role in the activation 19 of the Mn 3 O 4 -PMS system [31,32]. However, there are still limits on its aggressive 20 catalytic capacity and application performances, which limit its further application to 21 some extent. At present, the optimized methods include carbon-encapsulated [33], 22 incorporating a metal-organic skeleton [34], doping more active metal ions, etc. [35]. 23 Among them, doping metal ions such as Co, Ni, Zn, Cu, Fe, Cr, etc. are more feasible 24 due to their high catalytic activity and wide availability. Co has been considered to be 25 the most effective PMS activator where Co 2+ is transformed to Co 3+ and then to Co 2+ 26 regeneration [36,37]. Ren  were added under continuous rapid stirring. Finally, 5 mL, 10 mM of potassium 15 hydroxide solution was dropped into the solution and stirred vigorously for 10 min. 16 The final mixed solution was transferred to a Teflon stainless steel autoclave and

Catalytic activity of Fe 3 O 4 /Co-Mn 3 O 4 12
Batch experiments were conducted to study the activation of PMS by where C 0 (mg/L) and C t (mg/L) are the concentrations of OTC at initial and t min in 8 solution, respectively. 9 In this study, the pseudo-first-order model was used to describe the catalytic rate 10 of OTC by Fe 3 O 4 /Co-Mn 3 O 4 . The kinetic equation was expressed as follows [42]: where k obs is the catalytic rate constant (min -1 ). 13 The intermediate products of the OTC catalytic reaction were determined by 14 mass spectrometry. The mineralization of OTC was obtained by measuring the 15 removal rate of total organic carbon (TOC). 16 17 18 The microstructure characteristics of the as-prepared samples were well 19 displayed by SEM ( Fig. 1a-d) and TEM ( Fig. 1e- (Fig. 2c). There was broad absorption band from 3200 to 3600 cm -1 which 2 represented the stretching vibration of free surface water. The characteristic peak at could be ascribed to the bending vibration of hydroxyl combined with Mn atoms [46]. 10 The results of magnetometer analysis of Fe 3 25 The catalytic degradation of OTC was performed at different pH values. As 26 shown in Fig. 5a and b, the degradation efficiency of OTC was between 82.41% and , respectively, and HSO 5 had stronger oxidation capacity [47], so the 5 catalytic rate of OTC would decrease in alkaline conditions. Moreover, under high pH 6 conditions precipitation of Co (OH) 2 should also occur, which might lead to further 7 degradation of catalytic activity. But in general, these results suggested that   11 As shown in Fig. 5c and d, both the degradation efficiency and rate constant of  The isoelectric point of the material was 3.1 (Fig. S1), while the potential of 19 OTC aqueous solution after addition of 0.1 M NaNO 3 , NaCl, NaHCO 3 and NaHA was

The effect of initial pollutant concentration and salts
after FFA was added, which indicated that 1 O 2 was the main active substance. It was 1 also demonstrated that non-free radicals were involved in the degradation process of 2 OTC. carbonyl and amino bonds in OTC [62]. Then, P2 lost two more methyl groups, and 11 oxidized-dehydrogenate to get P3 (m/z 388.72) [63]. P3 was further removed from the 12 hydroxyl group, carbonyl group, double bond, and amino group, and the carbonyl 13 group was hydrogenated to produce P1. Round 3: P4 (m/z 455.12) was produced by 14 OTC through the loss of an OH group [64]. P4 lost part of its functional group, 15 oxidizes and then hydrogenated to form P1. Round 4: OTC first oxidized and lost part 16 of the functional group to form P5 (m/z 355.07). P5 then degraded to produce P1. In 17 addition, we analyzed the predictive and mutagenic toxicity of OTC degradation 18 intermediates (Fig. S2). In the predicted toxicity (Fig. S2a), although the 19 transformation products were relatively toxic, the toxicity of all the products except 20 P1 and P3 showed a decrease from the toxicity of mutagenesis (Fig. S2b). Therefore, 21 the toxicity of products after degradation was decreased.

22
XPS was employed to give insights into the chemical states and changes of the  (Fig. 7). The peaks at 638.1, 639.1, 640.2, and 641.8 eV corresponded to Mn 2+ , Mn 3+ , 25 Mn 4+ and satellite peaks, respectively (Fig. 7b) [31]. After the catalytic reaction, the  (Fig. 7c) [58]. After catalytic reaction, the 3 content of Co 2+ decreased from 57.49% to 39.07%, while that of Co 3+ increased from 4 42.51% to 60.93%. The decrease of Co 2+ content fully indicated that part of Co 2+ was 5 converted into Co 3+ , suggesting the occurrence of electron transfer within the metal. 6 At the same time, the binding energy of both Co 2+ (+1.32 eV) and Co 3+ (+1.29-1.5 eV) 7 increased after the reaction. As shown in Fig. 7d, the element Fe2p was divided into metals with different valence states [58,59]. The cyclic redox reactions in the metal 14 proved that the materials were not simply superimposed, but a synergistic catalytic 15 effect among the metals. 16 The removal efficiency and mineralization of OTC under different contact time 17 were shown in Fig. 8a the binding site on the OTC. Subsequently, the OTC was degraded directly by the 28 reactive complex through an abstract electron process [60]. The possible degradation 29 process of OTC was displayed in Fig. 9.   The catalytic performances of Fe 3 O 4 /Co-Mn 3 O 4 in actual water samples were also 21 conducted at different catalyst dosage and initial OTC concentration. Compared with 22 OTC removal in deionized water, high catalytic efficiency (about 98.8%) and rate 23 constant of OTC in actual water samples could be achieved at low catalyst dosage of 24 15 mg (Fig. 8d). At the same time, the removal efficiency was reached up to 76.12% 25 even at the OTC concentration of 50 mg/L (Fig. 8e). This fully proved that 26   the manuscript. All co-authors contributed to this work.

10
Ethics statement 11 No animal studies are presented in this manuscript. No human studies are 12 presented in this manuscript. No potentially identifiable human images or data are 13 presented in this study.
14 Consent for participation 15 Not applicable. 16 Consent for publication 17 Not applicable. 1 The authors declare that they have no known competing financial interests or 2 personal relationships that could have appeared to influence the work reported in this 3 paper.  The N2 adsorption-desorption isotherms of Fe3O4, Mn3O4, Co-Mn3O4 and Fe3O4/Co-Mn3O4 (a); XRD patterns of Fe3O4, Fe3O4/Co-Mn3O4 (b) and the illustration were XRD patterns of Mn3O4 and Co-Mn3O4; FTIR spectra of Fe3O4 and Fe3O4/Co-Mn3O4 before OTC degradation and the FTIR spectra of Fe3O4/Co-Mn3O4 after OTC degradation (c); Magnetization curves of Fe3O4 and Fe3O4/Co-Mn3O4 (d).

Figure 3
The comparison diagram of different components (a) and its kinetic tting curve (b); the optimization of different Fe3O4 contents (c) and their rate constant kobs (d).

Figure 4
The optimization of different Co contents (a) and their rate constant kobs (b); the (Ct/C0) -t curves of different dosing amounts (c) and their rate constant kobs (d).   XPS total absorption peak spectra before and after catalysis (a), and XPS absorption peak spectra of Mn (b), Co (c) and Fe (d) element before and after catalysis.

Figure 8
The removal e ciency of OTC and TOC (a); the reuse experiment (b); catalysis spectra (c) of Yellow River water, the changes of dosage (d) (the illustration showed the rate constant of Yellow River and deionized water at 15 mg), and concentration (e) (the illustration showed the rate constant of Yellow River and deionized water at 20 mg and 10 mg/L).

Figure 9
The possible degradation process of OTC.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. supplementarymaterial.docx