Mn 3 O 4 Catalysts for Advanced Oxidation of Phenolic Contaminants in Aqueous Solutions

: Water-soluble organic pollutants, such as phenolic compounds, have been exposed to environments globally. They have a signiﬁcant impact on groundwater and surface water quality. In this work, different Mn 3 O 4 catalysts were prepared for metal oxide activation of peroxymonosulfate (PMS) to remove the phenolic compound from the water environment. The as-prepared catalysts were characterized using thermogravimetric-differential thermal analysis (TG-DTA), powder X-ray diffraction (XRD), scanning electron microscopy (SEM), and Brunauer-Emmett-Teller (BET) surface area analysis. Furthermore, the effect of temperature and reusability of the as-prepared Mn 3 O 4 catalysts is also investigated. The Mn 3 O 4 nanoparticles (NPs) catalyst reveals an excellent performance for activating PMS to remove phenol compounds. Mn 3 O 4 NPs exhibits 96.057% efﬁciency in removing 25 ppm within 60 min. The kinetic analysis shows that Mn 3 O 4 NPs ﬁtted into pseudo-ﬁrst order kinetic model and exhibited relatively low energy activation of 42.6 kJ/mol. The reusability test of Mn 3 O 4 NPs displays exceptional stability with 84.29% efﬁciency after three-sequential cycles. The as-prepared Mn 3 O 4 NPs is proven suitable for phenolic remediation in aqueous solutions.


Introduction
Rapid economic growth has benefited socio-economic development, but has increased the risk of environmental issues [1]. One of the most severe environmental problems is water pollution. The limited availability of freshwater, which only accounts for 2.5% of the planet's water, faces water security threats that can affect 80% of the world's population [2]. The emerging contaminants (ECs), such as organic pollutants, pesticides, and pharmaceuticals, have been increased through the mismanagement of industrial wastewater. Watersoluble organic pollutants, mainly phenolic compounds, are particularly widespread, with approximately 3 million tonnes identified in the environment globally [3]. The phenolic effluents come mostly from coal conversion processes, petroleum refineries, phenolic resin manufacturing, herbicide manufacturing, pharmaceuticals, pulp and paper, and petrochemicals [3,4]. Phenolic compounds are extremely harmful to humans due to their acute toxicity, mutagenicity, and bio-recalcitrant characteristic [3,5].
Different biological, chemical, and physical processes have been applied to remove organic pollutants from an aqueous solution, such as adsorption, biodegradation, coagulationflocculation, ion exchange, photocatalytic, membrane electrodialysis and filtration, and solvent extraction. These strategies are incapable of completely removing phenolic compounds, require more extended time, and are not cost-effective [3,[6][7][8][9][10][11]. Advanced oxidation processes (AOPs) have been proven to be a promising alternative to water treatment due to their compelling oxidation potential for the mineralization of organic pollutants [12].

Mn 3 O 4 Catalysts Preparation
Mn 3 O 4 950 • C was synthesized from as-prepared spherical MnO 2 using the sonochemical method from Sankar et al., study [23]. Firstly, 0.8 g of potassium permanganate was dissolved in 50 mL of deionized water (DI water), then 10 mL of PEG200 was added to the aqueous solution. Afterward, the mixture was subjected to ultrasonication with a 20 kHz frequency for 15 min, where the reaction reached around 80 • C. The resulting brown precipitate was vacuum filtered using 0.45 µm Merck Millipore PVDF membrane and washed, firstly using DI water, and subsequently using ethanol. The obtained MnO 2 was dried at 80 • C for 8 h. Finally, the as-prepared MnO 2 was calcined at 950 • C. The as-prepared catalyst was denoted as Mn 3 O 4 950 • C.
Secondly, the Mn 3 O 4 NPs and octahedral catalysts were prepared using a hydrothermal method described by Li et al. [24], with some modifications. Firstly, Mn 3 O 4 NPs was prepared using 0.105 g of potassium permanganate precursor, dispersed into 30 mL of DI water. Subsequently, 15 mL of PEG200 was added to the solution. The mixture was mixed for 30 min at ambient temperature before being put into a Teflon-line stainless steel autoclave. The mixture then underwent a hydrothermal process for 8 h at 120 • C. The as-prepared products were collected using vacuum filtration and washed several times using DI water and ethanol before drying at 60 • C for 12 h. The as-prepared catalyst was denoted as Mn 3 O 4 -NPs.
Lastly, octahedral Mn 3 O 4 was prepared using a similar hydrothermal method with higher PEG200 concentrations. Five mol of potassium permanganate was dissolved into  30 mL of DI water with the addition of 30 mL of PEG200. Then, the as-prepared catalyst was denoted as octahedral Mn 3 O 4 .

Characterizations
The thermogravimetric analysis (TGA) (TGA/DSC1, Mettler Toledo, Columbus, OH, USA) was used to analyze the thermogravimetric profile of the MnO 2 . As-prepared catalysts were characterized for powder X-ray diffraction (XRD) using the Bruker-AXS D8 model (Bruker, Hannover, Germany) equipped with filtered Cu Kα radiation with a wavelength of 1.54178 Å and a scan range of 2θ of 5-70 • . The surface morphology of the samples was examined using the NEON 40EsB model (ZEISS, Oberkochen, Germany). The average length and diameter of the samples were calculated using the image processing program ImageJ from 50 individual structures. Micromeritics Tristar 3000 (Micromeritics, Norcross, GA, USA) was used to acquire Brunauer-Emmett-Teller (BET) specific surface area, pore volume, and average pore radius.

Phenolic Degradation Performance
The typical catalytic oxidation of phenol was conducted in a 500 mL reactor containing 0.2 g/L as-prepared catalysts and immersed in 25 ppm, 250 mL phenol solutions. The reactor was dipped in a water bath, which had an attached temperature controller and stirring rod with a rotational speed of 400 rpm. Subsequently, after adsorption-desorption equilibrium is achieved, 2 g/L OXONE ® was added to begin the oxidation process. At certain time intervals, 1 mL liquid solution was withdrawn from the reactor using a PVDF syringe filter of 0.45 µm. The liquid was then added to 0.5 mL methanol to prevent further reaction. The liquid concentrations were analyzed using high-performance liquid chromatography (HPLC), which is equipped with a UV detector at λ = 270 nm. The C-18 HPLC column was used with the mobile phase of 30% CH 3 CN and 70% ultrapure H 2 O with flowrate 1 mL/min. For chosen samples, total organic carbon (TOC) was defined applying a Shimadzu TOC-5000 CE analyzer.
The phenolic compound degradation efficiency was calculated using the following Equation (1): where C 0 and C e is the initial and remaining phenol concentrations, respectively. Furthermore, the kinetic analysis of the as-prepared samples were calculated using Langmuir-Hinshelwood (L-H) pseudo first-order kinetic equation as follow in Equation (2) [25]: where C 0 and C t are the initial and final phenol concentration, respectively, k is rate constant (min −1 ), and t is time (min). Additionally, the activation energy of the as-prepared catalysts was calculated at different temperatures (25-45 • C) using the Arrhenius equation. The Arrhenius equation, which correlated the relationship between the observed rate constant (k obs ) and temperature, was presented in the following Equation (3) [26,27]: where A is the pre-exponential factor which can be calculated from the intercept of the Arrhenius plot, E a. is activation energy (Kj.mol −1 ), T is temperature (K), and R is the gas constant (8.314 J/mol K). Moreover, the reusability test was done with the solid catalyst from the previous experimental run. Before the next experimental run, the solid catalyst was washed thoroughly with DI water and dried at 60 • C for 12 h.

Thermogravimetric Profile of MnO 2
The thermal stability and evolution profile of MnO 2 are presented in Figure 1. The initial weight loss of 4.3361% in the range of 30-200 • C corresponds to the evaporation of adsorbed water content on the surface of MnO 2 [28,29]. Moreover, observable two-phase weight losses were observed at 550 • C and 940 • C with 12.0075% and 16.2841% reduction, respectively. The first phase of weight loss at 550 • C is attributed to the evolution of MnO 2 to Mn 2 O 3 due to the oxygen release and reduction of Mn 4+ ions into Mn 3+ ions [30,31] The second phase of weight loss at 940 • C correlates with the sustained oxygen loss resulting in the conversion of Mn 2 O 3 to Mn 3 O 4 [30]. The TGA profile of Mn 3 O 4 concurrence with previous studies that reported the thermal stability of MnO 2 materials [30,32]. Furthermore, the transformation of as-prepared samples was examined further with XRD analysis. This TGA finding led to the choice of temperature for the hydrothermal synthesis of Mn 3 O 4 catalysts at 950 • C.
Moreover, the reusability test was done with the solid catalyst from the previous experimental run. Before the next experimental run, the solid catalyst was washed thoroughly with DI water and dried at 60 °C for 12 h.

Thermogravimetric Profile of MnO2
The thermal stability and evolution profile of MnO2 are presented in Figure 1. The initial weight loss of 4.3361% in the range of 30-200 °C corresponds to the evaporation of adsorbed water content on the surface of MnO2 [28,29]. Moreover, observable two-phase weight losses were observed at 550 °C and 940 °C with 12.0075% and 16.2841% reduction, respectively. The first phase of weight loss at 550 °C is attributed to the evolution of MnO2 to Mn2O3 due to the oxygen release and reduction of Mn 4+ ions into Mn 3+ ions [30,31] The second phase of weight loss at 940 °C correlates with the sustained oxygen loss resulting in the conversion of Mn2O3 to Mn3O4 [30]. The TGA profile of Mn3O4 concurrence with previous studies that reported the thermal stability of MnO2 materials [32,30]. Furthermore, the transformation of as-prepared samples was examined further with XRD analysis. This TGA finding led to the choice of temperature for the hydrothermal synthesis of Mn3O4 catalysts at 950 °C .

Crystallinity Structure, Morphology, and Surface Area Analysis
The crystallinity of as-prepared Mn3O4 is observed in Figure 2a [24] The broad diffraction of the as-prepared samples possesses nanostructure dimensions, as seen in previous work [24]. Subsequently, Figure 2c

Crystallinity Structure, Morphology, and Surface Area Analysis
The crystallinity of as-prepared Mn 3 O 4 is observed in Figure 2a [24] The broad diffraction of the as-prepared samples possesses nanostructure dimensions, as seen in previous work [24]. Subsequently, Figure 2c exhibited different peak intensities of XRD spectra for the as-prepared samples. The Mn 3 O 4 950 • C sample reveals the lowest intensity, followed by octahedral Mn 3 O 4 and Mn 3 O 4 NPs. The difference in peak intensity can be associated with different crystallite sizes and dislocation densities. The higher intensities correspond to larger crystallite sizes and better crystallinity. It is found that a larger crystallite size and higher crystallinity induced better degradation of organic compounds, as seen in a previous study [33]. Moreover, the determination of morphology of as-prepared Mn 3 O 4 is presented using SEM imaging. followed by octahedral Mn3O4 and Mn3O4 NPs. The difference in peak intensity can be associated with different crystallite sizes and dislocation densities. The higher intensities correspond to larger crystallite sizes and better crystallinity. It is found that a larger crystallite size and higher crystallinity induced better degradation of organic compounds, as seen in a previous study [33]. Moreover, the determination of morphology of as-prepared Mn3O4 is presented using SEM imaging.   Figure 3c) reveals more dispersed nanoparticles with octahedral shapes, and slight nanorod shapes can be observed. The length of octahedral Mn3O4 is 0.65 ± 0.1 µ m. Octahedron shapes could be formed utilizing accelerated reduction, and nucleation growth. However, the nanorod shapes could occur with excess PEG as a reducing agent and shape-directing agent [34].   (Figure 3c) reveals more dispersed nanoparticles with octahedral shapes, and slight nanorod shapes can be observed. The length of octahedral Mn 3 O 4 is 0.65 ± 0.1 µm. Octahedron shapes could be formed utilizing accelerated reduction, and nucleation growth. However, the nanorod shapes could occur with excess PEG as a reducing agent and shape-directing agent [34].  Figure 4 illustrates the morphological evolution of Mn3O4 from nanoplatelet to octahedral architecture. Firstly, Mn3O4 forms a nanoplatelet-like shape. Afterward, the Mn3O4 is agglomerated, forming a pre-octahedral shape, depicted in Figure 3b highlighted with the red circles. Then, the Mn3O4 formed dispersed octahedral shapes ( Figure 3c). The morphological evolution happened due to higher PEG200 concentrations. The polymer acts as a reducing agent and shape-directing agent [34]. The morphological evolution is followed by a self-assembly mechanism and then an Ostwald ripening mechanism [24]. The BET surface area, and N2 adsorption-desorption of each catalyst was analyzed to determine surface area, pore volume, and average pore radius. Table 1 lists the summary of the BET analysis. It reveals Mn3O4 950 °C possesses SBET 156.0 m 2 /g with a pore volume of 0.24 cm 3 /g and an average pore radius of 31.1 Å . The calculated surface area of Mn3O4 at 950 °C exhibited a higher BET surface area compared to other quasi-spherical Mn3O4 from previous studies, e.g., chemical leached Mn3O4 from manganese ore with a surface area of 9.32 m 2 /g [34], Mn3O4 synthesized using a precipitation method from potassium permanganate with a surface area of 21.2 m 2 /g [35], and solvothermal Mn3O4 from manganese(II)acetate tetrahydrate with a surface area of 131.49 m 2 /g [36]. Furthermore, different concentrations of the PEG200 alters the surface area of Mn3O4. Mn3O4 NPs exhibited higher surface area of 184.6 m 2 /g with an agglomerated platelet-like shape. With higher PEG200 concentrations, the Mn3O4 NPs transform from agglomerated nanoplatelets that resemble pre-octahedra into more polished octahedral shapes. Octahedral-shape controlled Mn3O4 was also exhibited in Li's study, where the Mn3O4 formation evolution is exhibited through different reaction times [24]. The as-prepared octahedral Mn3O4 possesses a lower surface area compared to Mn3O4 NPs. This is due to the smaller size of individual nanoplatelets of Mn3O4 NPs compared to octahedral shapes of octahedral Mn3O4. This phenomenon can be seen from previous transformation morphology studies of Mn3O4 [24,34]. The calculated SBET of octahedral Mn3O4 is 122.4 m 2 /g is proven to be higher than octahedral Mn3O4 from previous works, i.e., hydrothermal Mn3O4 from KMnO4 with SBET of 23 m 2 /g [37] and hydrothermal Mn3O4 from manganese(II) acetate with SBET of 57.7 m 2 /g [38]. Moreover, all three as-prepared catalysts observed micropore-type   (Figure 3c). The morphological evolution happened due to higher PEG200 concentrations. The polymer acts as a reducing agent and shape-directing agent [34]. The morphological evolution is followed by a self-assembly mechanism and then an Ostwald ripening mechanism [24].  Figure 4 illustrates the morphological evolution of Mn3O4 from nanoplatelet to octahedral architecture. Firstly, Mn3O4 forms a nanoplatelet-like shape. Afterward, the Mn3O4 is agglomerated, forming a pre-octahedral shape, depicted in Figure 3b highlighted with the red circles. Then, the Mn3O4 formed dispersed octahedral shapes (Figure 3c). The morphological evolution happened due to higher PEG200 concentrations. The polymer acts as a reducing agent and shape-directing agent [34]. The morphological evolution is followed by a self-assembly mechanism and then an Ostwald ripening mechanism [24]. The BET surface area, and N2 adsorption-desorption of each catalyst was analyzed to determine surface area, pore volume, and average pore radius. Table 1 lists the summary of the BET analysis. It reveals Mn3O4 950 °C possesses SBET 156.0 m 2 /g with a pore volume of 0.24 cm 3 /g and an average pore radius of 31.1 Å . The calculated surface area of Mn3O4 at 950 °C exhibited a higher BET surface area compared to other quasi-spherical Mn3O4 from previous studies, e.g., chemical leached Mn3O4 from manganese ore with a surface area of 9.32 m 2 /g [34], Mn3O4 synthesized using a precipitation method from potassium permanganate with a surface area of 21.2 m 2 /g [35], and solvothermal Mn3O4 from manganese(II)acetate tetrahydrate with a surface area of 131.49 m 2 /g [36]. Furthermore, different concentrations of the PEG200 alters the surface area of Mn3O4. Mn3O4 NPs exhibited higher surface area of 184.6 m 2 /g with an agglomerated platelet-like shape. With higher PEG200 concentrations, the Mn3O4 NPs transform from agglomerated nanoplatelets that resemble pre-octahedra into more polished octahedral shapes. Octahedral-shape controlled Mn3O4 was also exhibited in Li's study, where the Mn3O4 formation evolution is exhibited through different reaction times [24]. The as-prepared octahedral Mn3O4 possesses a lower surface area compared to Mn3O4 NPs. This is due to the smaller size of individual nanoplatelets of Mn3O4 NPs compared to octahedral shapes of octahedral Mn3O4. This phenomenon can be seen from previous transformation morphology studies of Mn3O4 [24,34]. The calculated SBET of octahedral Mn3O4 is 122.4 m 2 /g is proven to be higher than octahedral Mn3O4 from previous works, i.e., hydrothermal Mn3O4 from KMnO4 with SBET of 23 m 2 /g [37] and hydrothermal Mn3O4 from manganese(II) acetate with SBET of 57.7 m 2 /g [38]. Moreover, all three as-prepared catalysts observed micropore-type The BET surface area, and N 2 adsorption-desorption of each catalyst was analyzed to determine surface area, pore volume, and average pore radius. Table 1 lists the summary of the BET analysis. It reveals Mn 3 O 4 950 • C possesses S BET 156.0 m 2 /g with a pore volume of 0.24 cm 3 /g and an average pore radius of 31.1 Å. The calculated surface area of Mn 3 O 4 at 950 • C exhibited a higher BET surface area compared to other quasi-spherical Mn 3 O 4 from previous studies, e.g., chemical leached Mn 3 O 4 from manganese ore with a surface area of 9.32 m 2 /g [34], Mn 3 O 4 synthesized using a precipitation method from potassium permanganate with a surface area of 21.2 m 2 /g [35], and solvothermal Mn 3 O 4 from manganese(II)acetate tetrahydrate with a surface area of 131.49 m 2 /g [36]. Furthermore, different concentrations of the PEG200 alters the surface area of Mn 3 O 4 . Mn 3 O 4 NPs exhibited higher surface area of 184.6 m 2 /g with an agglomerated platelet-like shape. With higher PEG200 concentrations, the Mn 3 O 4 NPs transform from agglomerated nanoplatelets that resemble pre-octahedra into more polished octahedral shapes. Octahedral-shape controlled Mn 3 O 4 was also exhibited in Li's study, where the Mn 3 O 4 formation evolution is exhibited through different reaction times [24]. The as-prepared octahedral Mn 3 O 4 possesses a lower surface area compared to Mn 3 O 4 NPs. This is due to the smaller size of individual nanoplatelets of Mn 3 O 4 NPs compared to octahedral shapes of octahedral Mn 3 O 4 . This phenomenon can be seen from previous transformation morphology studies of Mn 3 O 4 [24,34]. The calculated S BET of octahedral Mn 3 O 4 is 122.4 m 2 /g is proven to be higher than octahedral Mn 3 O 4 from previous works, i.e., hydrothermal Mn 3 O 4 from KMnO 4 with S BET of 23 m 2 /g [37] and hydrothermal Mn 3 O 4 from manganese(II) acetate with S BET of 57.7 m 2 /g [38]. Moreover, all three as-prepared catalysts observed micropore-type pore size. This classification is determined by IUPAC classification with pore diameter of <2 nm, 2-50 nm, and > 50 nm classified as micropore, mesopore and macropore, respectively.

Phenolic Degradation Performance
The catalytic activity of the as-prepared Mn 3 O 4 catalysts is evaluated from phenolic degradation using PMS presented in Figure 5. The initial PMS degradation without transition metal activation is performed to identify the contribution of PMS to phenolic degradation. It is found that PMS without activation degraded 2.465% phenolic compound, which has a negligible effect on the overall performance. This is because, with PMS alone, the sulfate radicals cannot be produced for phenolic oxidation [39]. Similar PMS without activation performance was also reported from previous work. Liu and Huang reported the role of PMS without activation is negligible, with only < 5% oxidation for the phenolic compound [40]. Furthermore, the preliminary performance of Mn 3 O 4 without the presence of PMS is investigated using a Mn 3 O 4 NPs sample. It is shown that the adsorption performance of Mn 3 O 4 adsorbed 9.242% phenolic compound within 60 min. The low adsorption capacity of Mn 3 O 4 is also recorded in previous studies. This occurs because phenol molecules easily penetrate through pores larger than 1 nm, while most NPs have pores smaller than this. Wang et al. reported no to little adsorption of phenol with only < 5% adsorption using Mn 3 O 4 2D nanosheet after 2 h reaction time [5]. The higher adsorption rate we achieved in comparison to Wang et al. is due to the as-prepared Mn 3 O 4 having a higher calculated surface area than in their work (65.2 m 2 /g). The two best-performing catalysts underwent investigation of different temperatures toward phenolic compound removal. Figure 6a shows octahedral Mn3O4 phenol removal performance at 25, 35, and 45 °C reaction temperatures. Moreover, the Langmuir-Hilselwood pseudo-first order kinetic model is fitted to determine the reaction rate [42]. Table 2    NPs possess the most surface area and pore volume of the three catalysts, and these are the active sites where PMS activation occurs [12]. Additionally, octahedral Mn 3 O 4 has the second-best oxidative degradation performance because it has a higher pore volume with a tighter pore radius, and therefore more active sites, compared to Mn 3 O 4 950 • C. The greater availability of active sites for PMS activation is attributed to the fast-generating rate of sulfate radicals [39,41].
The two best-performing catalysts underwent investigation of different temperatures toward phenolic compound removal. Figure 6a shows octahedral Mn 3 O 4 phenol removal performance at 25, 35, and 45 • C reaction temperatures. Moreover, the Langmuir-Hilselwood pseudo-first order kinetic model is fitted to determine the reaction rate [42]. Table 2 shows all the constant rates and the correlation coefficient of 0.99. The degradation result of octahedral Mn 3 O 4 at 45, 35, and 25 • C all achieved 100% removal within 40, 120, and 190 min with k obs of 0.128, 0.073, and 0.030 min −1 , respectively. Additionally, Figure 6b shows Mn 3 O 4 NPs results at 45, 35, and 25 • C achieved 100%, 99.408%, and 96.057% removal within 40, 60, and 60 min with k obs of 0.186, 0.073, and 0.043 min −1 , respectively. The result shows that higher temperature ensued better oxidation of phenol compounds. Higher temperature stimulates the chemical bond breakage and accelerates the decomposition of the persulfate, which leads to increased phenol removal [43][44][45]. TOC removal in Mn 3 O 4 NPs/PMS systems was also measured and the results showed that roughly 78% of TOC removal was obtained for Mn 3 O 4 NPs/PMS, at 190 min. The two best-performing catalysts underwent investigation of different temperatures toward phenolic compound removal. Figure 6a shows octahedral Mn3O4 phenol removal performance at 25, 35, and 45 °C reaction temperatures. Moreover, the Langmuir-Hilselwood pseudo-first order kinetic model is fitted to determine the reaction rate [42]. Table 2 shows all the constant rates and the correlation coefficient of 0.99. The degradation result of octahedral Mn3O4 at 45, 35, and 25 °C all achieved 100% removal within 40, 120, and 190 min with kobs of 0.128, 0.073, and 0.030 min −1 , respectively. Additionally, Figure  6b shows Mn3O4 NPs results at 45, 35, and 25 °C achieved 100%, 99.408%, and 96.057% removal within 40, 60, and 60 min with kobs of 0.186, 0.073, and 0.043 min −1 , respectively. The result shows that higher temperature ensued better oxidation of phenol compounds. Higher temperature stimulates the chemical bond breakage and accelerates the decomposition of the persulfate, which leads to increased phenol removal [43][44][45]. TOC removal in Mn3O4 NPs/PMS systems was also measured and the results showed that roughly 78% of TOC removal was obtained for Mn3O4 NPs/PMS, at 190 min.  The correlation between temperature and reaction rate is further studied using the Arrhenius equation plot. Figure 7a,b show the Arrhenius plot for the as-prepared catalysts with a correlation coefficient of 0.99. It is found that octahedral Mn 3 O 4 has an energy Water 2022, 14, 2124 9 of 13 activation of 71.23 kJ/mol, which is much higher than Mn 3 O 4 NPs with E a of 42.6 kJ/mol. The energy activation of both catalysts is higher than diffusion-controlled reaction, which is in the range of 10-13 kJ/mol. This higher value indicates that the reaction rate is predominantly an intrinsic chemical reaction on the surface of oxide rather than a rate of mass transfer [46].

Mn3O4
71. 23 35 0.073 0.99 45 0.128 0.99 The correlation between temperature and reaction rate is further studied using the Arrhenius equation plot. Figure 7a,b show the Arrhenius plot for the as-prepared catalysts with a correlation coefficient of 0.99. It is found that octahedral Mn3O4 has an energy activation of 71.23 kJ/mol, which is much higher than Mn3O4 NPs with Ea of 42.6 kJ/mol. The energy activation of both catalysts is higher than diffusion-controlled reaction, which is in the range of 10-13 kJ/mol. This higher value indicates that the reaction rate is predominantly an intrinsic chemical reaction on the surface of oxide rather than a rate of mass transfer [46]. In addition, the comparison of PMS activation using several related catalysts for phenolic compound degradation is presented in Table 3. The as-prepared Mn3O4 NPs catalyst has relatively lower energy activation than several catalysts from previous works. Activation energy gives an insight into the minimum energy required for a chemical reaction to occur. The degradation of catalysts with lower activation energy starts faster compared to catalysts with higher activation energy [47].  In addition, the comparison of PMS activation using several related catalysts for phenolic compound degradation is presented in Table 3. The as-prepared Mn 3 O 4 NPs catalyst has relatively lower energy activation than several catalysts from previous works. Activation energy gives an insight into the minimum energy required for a chemical reaction to occur. The degradation of catalysts with lower activation energy starts faster compared to catalysts with higher activation energy [47]. NPs maintained an efficiency of 84.29% within 60 min after the third recycling, with only an 11.81% decline. On the other hand, octahedral Mn 3 O 4 exhibited a 9.25% decrease in degradation performance with 87.35% efficiency within 150 min reaction time after the third recycle. The decline in efficiency can be attributed to the surface deactivation of the catalysts with intermediates on their surface [51]. However, the results of both catalysts show a considerably good performance after the third cycle since no additional purification is performed, except for simply washing them with water. Nevertheless, the best way to remove the intermediates from the catalyst surface is to calcinate and wash with ammonia solution, as Sun et al. reported [52]. NPs maintained an efficiency of 84.29% within 60 min after the third recycling, with only an 11.81% decline. On the other hand, octahedral Mn3O4 exhibited a 9.25% decrease in degradation performance with 87.35% efficiency within 150 min reaction time after the third recycle. The decline in efficiency can be attributed to the surface deactivation of the catalysts with intermediates on their surface [51]. However, the results of both catalysts show a considerably good performance after the third cycle since no additional purification is performed, except for simply washing them with water. Nevertheless, the best way to remove the intermediates from the catalyst surface is to calcinate and wash with ammonia solution, as Sun et al. reported [52].

Proposed Mechanism of PMS Activation by Mn3O4 Catalyst
The oxidation of organic compounds started when PMS contacted the active sites on the surface of Mn4. Then the active sites transferred a donor electron via a reduction-oxidation reaction, both Mn(III) and Mn(IV). Mn(IV) was reduced to Mn(III) and afterward oxidized to Mn(IV) using HSO5¯. Subsequently, Mn(IV) activated PMS to produce SO4 •− , then released into the bulk solution and interacted with H2O and OH to produce • OH.

Proposed Mechanism of PMS Activation by Mn 3 O 4 Catalyst
The oxidation of organic compounds started when PMS contacted the active sites on the surface of Mn 4 . Then the active sites transferred a donor electron via a reductionoxidation reaction, both Mn(III) and Mn(IV). Mn(IV) was reduced to Mn(III) and afterward oxidized to Mn(IV) using HSO 5 −. Subsequently, Mn(IV) activated PMS to produce SO 4 •− , then released into the bulk solution and interacted with H 2 O and OH to produce • OH. The generated SO 4 •− and • OH would target phenol and produce numerous intermediates until the phenol is broken down into CO 2 and H 2 O. The full step by step mechanism is illustrated in the following Equation (4) and Figure 9. The generated SO4 • − and • OH would target phenol and produce numerous intermediates until the phenol is broken down into CO2 and H2O. The full step by step mechanism is illustrated in the following Equation (4) and Figure 9.

Conclusions
Three different Mn3O4 catalysts with different crystallinity, morphology, and specific surface areas were successfully synthesized using a simple hydrothermal method. The

Conclusions
Three different Mn 3 O 4 catalysts with different crystallinity, morphology, and specific surface areas were successfully synthesized using a simple hydrothermal method. The Mn 3 O 4 NPs exhibit excellent performance in activating PMS for phenol removal with high efficiency and relatively low energy activation compared to other related catalysts.