Metal-Loaded Mesoporous MCM-41 for the Catalytic Wet Peroxide Oxidation (CWPO) of Acetaminophen

: MCM-41 based catalysts (molar ratio Si/Al = 40) were prepared by a hydrothermal route, modiﬁed by ionic exchange with different metals (Cu, Cr, Fe and Zn) and ﬁnally calcined at 550 ◦ C. The catalysts were fully characterized by different techniques that conﬁrmed the formation of oxides of the different metals on the surfaces of all materials. Low-angle X-ray diffraction (XRD) analyses showed that calcination resulted in the incorporation of metallic Zn, Fe and Cr in the framework of MCM-41, while in the case of Cu, thin layers of CuO were formed on the surface of MCM-41. The solids obtained were tested in the catalytic wet peroxide oxidation (CWPO) of acetaminophen at different temperatures (25–55 ◦ C). The activity followed the order: > Zn/MCM-41. The increase of the reaction temperature improved the performance and activity of Cr/MCM-41 and Fe/MCM-41 catalysts, which achieved complete conversion of acetaminophen in short reaction times (15 min in the case of Cr/MCM-41). Fe/MCM-41 and Cr/MCM-41 were submitted to long-term experiments, being the Fe/MCM-41 catalyst the most stable with a very low metal leaching. The leaching results were better than those previously reported in the literature, conﬁrming the high stability of Fe/MCM-41 catalysts synthesized in this study. HCl solution. The samples were taken at different times in maximum 2 h and they were analyzed by high-performance liquid chromatography (HPCL) in order to obtain the adsorption capacities of catalysts. The results show that adsorption equilibrium was obtained after 1 h with very low adsorption capacities.


Introduction
In recent years, global concern about the presence of pharmaceutical compounds in water has significantly increased. A wide variety of pharmaceutical compounds, such as analgesics, anti-inflammatory, antibiotics, contrast or antiepileptic agents, has recently been detected in many water resources [1]. The pharmaceutical compounds found in the environment come mainly from the elimination of these molecules and their metabolites by humans or animals under therapeutic treatment and/or and inadequate treatment of manufacturing effluents [2]. Because of their stabilities and sometimes recalcitrant behavior, they are not completely removed/degraded by conventional wastewater treatment techniques [3][4][5]. Their concentrations in water, although depending on the specific site and compound, are generally low, in the range of micrograms to nanograms per liter [3][4][5], although in the case of wastewaters coming from manufacturing plant effluents they could The main objective of this work is the preparation of different mesoporous materials containing different transition metals (Cu, Cr, Fe and Zn) using the non-calcined form of MCM-41 in order to increase the metal loading of the final catalysts. The catalysts were tested in the CWPO of acetaminophen and the proper structure-performance relationships were detailed and discussed. Figure 1 represents the X-ray diffraction (XRD) patterns at low and large angles of MCM-41 and the different catalysts. The XRD pattern of the original MCM-41 sample shows the presence of an intense peak at a 2θ value close to 2.5 • , besides to lower intensity peaks at 2θ values around to 3.7 and 4.3 • , characteristic of the reflection of (100), (110) and (200) planes of this mesoporous solid aluminosilicate. The absence of peaks at higher angles is because MCM-41 consists of amorphous silica, and thus it has no crystallinity at the atomic level [36] but only a high pore ordering. The different catalysts show clear differences in the XRD patterns. Cu/MCM-41 catalyst retains a high intensity (100) peak, besides two lower intensity peaks (110) and (200) at 2θ ≈ 3.72 and 4.30 • , respectively. This confirms that this sample maintains the hexagonal structures of the MCM-41 with a high degree of structural ordering [37]. The XRD patterns of the rest of the catalysts, Cr/MCM-41, Fe/MCM-41 and Zn/MCM-41, show much lower intensity of (100) peak and are displaced to lower diffraction angles. This can be explained by the swelling of the hexagonal mesh due to the incorporation of the metal in the MCM-41 framework (see Table 1) in agreement with previous observations [30,38]. The modification of the mesh parameters confirms that the calcination at 550 • C leads to the incorporation of the metal in the inner of the MCM-41 framework. The incorporation of a metal into the framework of the MCM-41 causes an increase in the mesh parameter depending on the size of the radius of the metal introduced in its structure [30,39]. According to the calculated values of a 0 , it is shown that all the modified materials have higher values compared to the parent MCM-41, except Cu/MCM-41 catalyst, which has an almost similar value compared to the parent MCM-41. In the large angles, all the catalysts presented new phases corresponding to the oxides of the metals added in the ion exchange process on the surface of MCM-41. Hence, the peaks at 2θ values of 35.5, 38.7, 42.3, 48,7 and 52.4 • correspond to the (-111), (111), (200), (-202) and (211) reflection phases of CuO respectively [38,40] [43,44]. This confirms that the calcination does not only lead to the incorporation of the metal into the MCM-41 framework but also forms metal oxides phases on its surface. Cu/MCM-41 catalyst has the higher intensity peaks in the region 30-60 • probably due to the formation of higher crystalline CuO outside the MCM-41 framework. This result, besides the low significant modification of the a o parameter, suggests than in the case of Cu, most of the metal was not incorporated into the Al-MCM-41 framework but deposited on the surface in CuO form. In the case of the Zn exchange, no ZnO phase was observed for the Zn/MCM-41 sample, suggesting this metal was deposited in the form of well-dispersed species with small size [36].  The nitrogen adsorption-desorption isotherms at 77 K of the the pattern MCM-41 are represented in Figure S1. All isotherms a istic of mesoporous materials with also a significant presence of m These isotherms are characterized by three steps; the first step is ca pressures (P/P0 < 0.25) which are characteristic of monolayer adsor micropore walls. The second step takes place at relative pressure  The ion metal exchanged with Cetyltrimethylammonium(CTA + ) in MCM-41 before calcination increases the incorporation inside the material [45][46][47] which means more metal oxide, the amount of the metal in each catalyst was quantified by X-ray fluorescence (XRF) being summarized in Table 1 The nitrogen adsorption-desorption isotherms at 77 K of the different catalysts and the pattern MCM-41 are represented in Figure S1. All isotherms are of type IV characteristic of mesoporous materials with also a significant presence of microporosity [29,30,48]. These isotherms are characterized by three steps; the first step is carried out at low relative pressures (P/P 0 < 0.25) which are characteristic of monolayer adsorption of nitrogen on the micropore walls. The second step takes place at relative pressures between 0.25 < P/P 0 < 0.4 which is characterized by a sharp increase in nitrogen adsorption. This adsorption behavior is due to capillary condensation within the uniform micropores. The third step starts from P/P 0 > 0.4 which corresponds to the gradual increase in volume, due to the multilayer adsorption on the outer surface of the mesoporous materials. Table 1 shows the textural properties of obtained materials. It can be seen that the exchange of MCM-41 with the different metals resulted in a reduction of the porosity of the resulting catalysts [29,30]. This is strongly related to the particle blocking of the internal porosity by the incorporation of the metallic species. The increase of the pore diameter compared to the parent material is due to the incorporation of the incorporation of the metals to the structure during the heat treatment. Fe/MCM-41 exhibited the highest pore diameter suggesting a significant incorporation of this metal into the framework of the MCM-41, while Cr/MCM-41 exhibited a slight modification (in agreement with the amount of metal quantified by XRF). The catalyst Cu/MCM-41 has the higher surface area compared to the other modified samples probably due to its low metal content (see Table 1). This result confirms that CuO was deposited in the wider mesopores of MCM-41, which is in agreement with previously published works [48].

Characterization of the Catalyst
The study of the thermal behavior of the materials obtained makes it possible to determine the mass loss of each sample with increasing temperature, but also provides information about the formation mechanism of metal-loaded-MCM-41. Figure S2 shows the thermogravimetric analysis (TGA) curves under nitrogen atmosphere (inert atmosphere to avoid weight gaining caused by the presence of oxygen in air atmosphere) of as-synthesized MCM-41 (parent material) and its counterparts modified by Cr, Fe, Zn and Cu, prior to calcination. All materials presented two masses losses, the first mass loss occurs between 25 and 120 • C and it is associated to the loss of physiosorbed water. The second mass loss takes place at temperatures between 120 and 700 • C and corresponds to different stages of degradation of the surfactant CTA + .
Infrared spectroscopy was used in order to determine qualitatively the different groups of the catalysts before and after calcination ( Figure S3). After calcination at 550 • C a wide band was formed at 3387 cm −1 and another band appears at 1635 cm −1 ( Figure S3b). These bands are due to the presence of the silanol groups and also to the vibrations of physiosorbed water. This also confirms that calcination not only removes the surfactant inside the pores of MCM-41, as indicates the disappearance of the bands at 2919 and 2848 cm −1 , but also leads to the formation of higher amounts of silanol groups compared to non-calcined materials. The bands at 1221, 1032 and 786 cm −1 for the non-calcined materials have been moved to 1238, 1048 and 794 cm −1 at the higher wavenumbers, respectively. This behavior is strongly linked to the incorporation of transition metals into the framework of MCM-41 due to the calcination treatment [49].
Ultraviolet-visible (UV-vis) spectroscopy was used in order to determine the coordination nature of the metallic species supported on the surface of MCM-41. As shown in Figure S4, the bare mesoporous silica only shows peaks at wavelengths in the 200-350 nm range corresponding to UV absorption of SiO 2 [50]. The introduction of Cu, Zn, Fe and Cr in MCM-41 results in significant changes in the corresponding UV-Vis spectra ( Figure 2). The catalyst Cu/MCM-41 displays two clear bands, an intense band at 219 nm and a wide band between 300-800 nm that are mainly due to the presence of Cu 2+ in form of CuO NPs [51,52]. Zn/MCM-41 catalyst is characterized by two bands, the first is located at 257 nm and the second and wide band located at 305 nm, which can be assigned to encapsulated ZnO NPs [53,54]. In general, the band below 220 nm confirms the incorporation of Zn into the framework of MCM-41 and is associated with the charge transfer transitions of the zinc species with the oxygen of the framework [55]. The UV spectrum of the catalyst Cr/MCM-41 is very similar to that obtained in the literature in which it clearly shows the existence of Cr 2 O 3 [56]. The catalyst Fe/MCM-41 presents a strong band at 215 nm associated with another band at 268 nm, which is assigned to the charge transfer between the Fe 3+ and O 2− atoms (Fe-O-Si) in the framework of MCM-41 [57]. In addition to these bands, a shoulder appeared around 330 and 500 nm, indicating the presence of either extra-framework iron or iron oxide particles [58]. The investigation of the electronic states of MCM-41 doped by different metals was carried out by X-ray photoelectron spectroscopy (XPS). Figure 3 represents the Cr, Cu, Zn and Fe 2p XPS signal of the correspondent catalysts. XPS spectra of Cr 2p showed the Cr 2p1/2 and Cr 2p3/2 peaks on Cr/MCM-41 were located at 586.3 and 576.7. eV, which are characteristic of Cr2O3 species [59,60]. The O1s XPS peak ( Figure S5 of the Supplementary DATA) has a peak at 532 eV, which corresponds to the oxygen of the Cr2O3 network [53] and also to the Si-O bond and explains a close interaction between oxygen and silicon [61]. The XPS spectra of Cu 2p of the catalyst Cu/MCM-41 have two Cu peaks, belonging respectively to Cu 2p3/2 at around 933 eV and Cu 2p1/2 at 953 eV, suggesting the presence of Cu 2+ species in form of CuO [62,63]. The Zn 2p peaks for the catalyst Zn/MCM-41 show the Zn 2p1/2 and Zn 2p3/2, contributions, with peaks at 1045.8 eV and 1022.6 eV ascribed to Zn 2+ species [64]. The value of Zn 2p1/2 -Zn 2p3/2 is approximately 23.12 eV corresponding to the typical splitting value of ZnO [65,66]. The Fe 2p XPS spectra of Fe/MCM-41 exhibited a strong peak at 711.0 eV with a lower intensity satellite peak at 719.2 eV assigned to Fe 2p3/2 signal. The weak satellite peak of around 719.52 eV is characteristic of Fe 3+ in Fe2O3 [67]. The binding energy of Fe 2p1/2 is around 724.84 eV associated with other satellite peak at 732.01 eV [68]. The peaks at around 711.01 eV and 724.84 eV, indicating Fe components were mainly in trivalent status in Fe-MCM-41 [69]. The XPS O1s peak of all the catalysts is very similar (see Figure S5). A first peak, between 533.0-533.2 eV, corresponds to the oxygen of SiO2 [70]. The other peak, located at around 532.0-530.0 eV, is attributed to T-OH (T = Si, Al or Metal), T-O-Si (T = Metal) and T-O (T = Metal) [70][71][72]. These peaks were almost masked by the intense and wide oxygen signal from SiO2 [73]. The investigation of the electronic states of MCM-41 doped by different metals was carried out by X-ray photoelectron spectroscopy (XPS). Figure 3 represents the Cr, Cu, Zn and Fe 2p XPS signal of the correspondent catalysts. XPS spectra of Cr 2p showed the Cr 2p 1/2 and Cr 2p 3/2 peaks on Cr/MCM-41 were located at 586.3 and 576.7 eV, which are characteristic of Cr 2 O 3 species [59,60]. The O1s XPS peak ( Figure S5 of the Supplementary DATA) has a peak at 532 eV, which corresponds to the oxygen of the Cr 2 O 3 network [53] and also to the Si-O bond and explains a close interaction between oxygen and silicon [61]. The XPS spectra of Cu 2p of the catalyst Cu/MCM-41 have two Cu peaks, belonging respectively to Cu 2p 3/2 at around 933 eV and Cu 2p 1/2 at 953 eV, suggesting the presence of Cu 2+ species in form of CuO [62,63]. The Zn 2p peaks for the catalyst Zn/MCM-41 show the Zn 2p 1/2 and Zn 2p 3/2 , contributions, with peaks at 1045.8 eV and 1022.6 eV ascribed to Zn 2+ species [64]. The value of Zn 2p 1/2 -Zn 2p 3/2 is approximately 23.12 eV corresponding to the typical splitting value of ZnO [65,66]. The Fe 2p XPS spectra of Fe/MCM-41 exhibited a strong peak at 711.0 eV with a lower intensity satellite peak at 719.2 eV assigned to Fe 2p 3/2 signal. The weak satellite peak of around 719.52 eV is characteristic of Fe 3+ in Fe 2 O 3 [67]. The binding energy of Fe 2p 1/2 is around 724.84 eV associated with other satellite peak at 732.01 eV [68]. The peaks at around 711.01 eV and 724.84 eV, indicating Fe components were mainly in trivalent status in Fe-MCM-41 [69]. The XPS O1s peak of all the catalysts is very similar (see Figure S5). A first peak, between 533.0-533.2 eV, corresponds to the oxygen of SiO 2 [70]. The other peak, located at around 532.0-530.0 eV, is attributed to T-OH (T = Si, Al or Metal), T-O-Si (T = Metal) and T-O (T = Metal) [70][71][72]. These peaks were almost masked by the intense and wide oxygen signal from SiO 2 [73].   Figure 4a,b represent the ACE concentration and H2O2 conversion upon reaction time when using the different catalysts. It should be remarked that the ACE adsorption (see Figure S6) was negligible in all th siloxy e cases. Despite the well-developed surface area of the MCM-41 sample, it shows an almost negligible adsorption capacity of acetaminophen. This is mainly due to the hydrophilic nature of MCM-41, and it is in agreement with previous studies [74]. All the tests were performed at pH = 3, which is generally accepted as the optimal pH to perform CWPO reactions. At lower pH values the formation of hydrated metal complexes results in a lower production of hydroxyl radicals. Furthermore, at very acidic conditions a higher number of protons is free in the medium to react with hydroxyl radicals, which decreases the overall efficiency of the process. On the other hand, pH > 4 favors the formation of metal oxyhydroxides on the catalyst's surface, which inhibits both the production of hydroxyl radicals and the regeneration of the metal ions. ACE conversion varies in the following sequence Cr/MCM-41 >> Fe/MCM-41 > Cu/MCM-41 >> Zn/MCM-41. The ACE conversion was completed in less than 60 min using Cr/MCM-41 while the Fe/MCM-41 catalyst needs four times higher reaction time to achieve complete ACE conversion. The low initial acetaminophen concentration (5 mg·L −1 ) makes difficult the identification and accurate quantification of the reaction intermediates. No additional aromatic compounds were observed in HPLC chromatograms. However, short-chain organic acids, such as, oxalic, acetic and formic acids, were detected using ionic chromatography. The difference between the ACE and H2O2 final conversions (at 4 h) when using Fe/MCM-41 catalyst should be ascribed to the presence of intermediate oxidation byproducts, suggesting an incomplete mineralization of ACE. In contrast, Zn/MCM-41 catalysts showed very little ACE conversion, hardly 10% after 240 min of reaction. This activity can be linked to several factors such as the nature of the metal oxides, their crystallinity, their metal particle sizes and/or their specific surface [75,76]. The  Figure 4a,b represent the ACE concentration and H 2 O 2 conversion upon reaction time when using the different catalysts. It should be remarked that the ACE adsorption (see Figure S6) was negligible in all th siloxy e cases. Despite the well-developed surface area of the MCM-41 sample, it shows an almost negligible adsorption capacity of acetaminophen. This is mainly due to the hydrophilic nature of MCM-41, and it is in agreement with previous studies [74]. All the tests were performed at pH = 3, which is generally accepted as the optimal pH to perform CWPO reactions. At lower pH values the formation of hydrated metal complexes results in a lower production of hydroxyl radicals. Furthermore, at very acidic conditions a higher number of protons is free in the medium to react with hydroxyl radicals, which decreases the overall efficiency of the process. On the other hand, pH > 4 favors the formation of metal oxyhydroxides on the catalyst's surface, which inhibits both the production of hydroxyl radicals and the regeneration of the metal ions. ACE conversion varies in the following sequence Cr/MCM-41 >> Fe/MCM-41 > Cu/MCM-41 >> Zn/MCM-41. The ACE conversion was completed in less than 60 min using Cr/MCM-41 while the Fe/MCM-41 catalyst needs four times higher reaction time to achieve complete ACE conversion. The low initial acetaminophen concentration (5 mg·L −1 ) makes difficult the identification and accurate quantification of the reaction intermediates. No additional aromatic compounds were observed in HPLC chromatograms. However, short-chain organic acids, such as, oxalic, acetic and formic acids, were detected using ionic chromatography. The difference between the ACE and H 2 O 2 final conversions (at 4 h) when using Fe/MCM-41 catalyst should be ascribed to the presence of intermediate oxidation byproducts, suggesting an incomplete mineralization of ACE. In contrast, Zn/MCM-41 catalysts showed very little ACE conversion, hardly 10% after 240 min of reaction. This activity can be linked to several factors such as the nature of the metal oxides, their crystallinity, their metal particle sizes and/or their specific surface [75,76]. The results illustrated in Figure 4 are not in agreement with the trend observed for the microporosity of the samples, suggesting that the catalytic activity of the catalysts is not strongly affected by the surface, as previously reported in other studies [20]. 021, 11, x FOR PEER REVIEW results illustrated in Figure 4 are not in agreement with the trend observed for croporosity of the samples, suggesting that the catalytic activity of the catalysts strongly affected by the surface, as previously reported in other studies [20].  (Table 1). Moreover, these samples have not been able to d pose H2O2, which justifies their lower activity in CWPO reaction. Fe/MCM-Cr/MCM-41 catalysts have higher metal contents. Several studies have shown t various oxide metals including magnetite [34,77,78] are effective heterogeneous catalysts which are in agreement with our results.

Catalytic Test
The best catalysts, Fe/MCM-41 and Cr/MCM-41, were used in the rest of the Figure 5 represents the evolution of ACE concentration and H2O2 conversion with r time when using different loads of Fe/MCM-41 and Cr/MCM-41 catalysts. The m catalyst was varied between 0.5 and 2 g/L while keeping the temperature (25 °C) a ≈ 3 of solution constant. The load of catalyst plays a very important role on the con of ACE that increases rapidly with increasing mass of the catalyst. A total conver ACE was obtained with Cr/MCM-41 catalyst at reaction time lower than 30 min using catalyst loads of 1 or 2 g/L. This is a consequence of the presence of more acti when using higher loads of catalyst. Another explanation also linked to the size chromium oxides which probably have a smaller size compared to the iron Cr/MCM-41 has a larger surface area than Fe/MCM-41, which may be due to the p of large particles of iron oxide inside the pores subsequently leading to a decrease number of sites, and thus a tortuous diffusion of reagents. Figure 5c,d shows H2O2 conversion using both catalysts. In the case of Cr/M higher conversions of H2O2 lead to higher and faster degradation of ACE. It sho noted that the reaction between H2O2 and ACE without catalyst does not lead to an radation which has been confirmed by Carrasco-Díaz et al. [20]. It has been show the total degradation of ACE requires approximately 45-55% of H2O2 conversion The low activity of Cu/MCM-41 and Zn/MCM-41 catalysts was probably linked to their lower metal content (Table 1). Moreover, these samples have not been able to decompose H 2 O 2 , which justifies their lower activity in CWPO reaction. Fe/MCM-41 and Cr/MCM-41 catalysts have higher metal contents. Several studies have shown that the various oxide metals including magnetite [34,77,78] are effective heterogeneous Fenton catalysts which are in agreement with our results.
The best catalysts, Fe/MCM-41 and Cr/MCM-41, were used in the rest of the work. Figure 5 represents the evolution of ACE concentration and H 2 O 2 conversion with reaction time when using different loads of Fe/MCM-41 and Cr/MCM-41 catalysts. The mass of catalyst was varied between 0.5 and 2 g/L while keeping the temperature (25 • C) and pH ≈ 3 of solution constant. The load of catalyst plays a very important role on the conversion of ACE that increases rapidly with increasing mass of the catalyst. A total conversion of ACE was obtained with Cr/MCM-41 catalyst at reaction time lower than 30 min when using catalyst loads of 1 or 2 g/L. This is a consequence of the presence of more active sites when using higher loads of catalyst. Another explanation also linked to the size of the chromium oxides which probably have a smaller size compared to the iron oxides, Cr/MCM-41 has a larger surface area than Fe/MCM-41, which may be due to the presence of large particles of iron oxide inside the pores subsequently leading to a decrease in the number of sites, and thus a tortuous diffusion of reagents. Figure 5c, [79], which are considered as Fenton-active sites for the degradation of ACE. The H 2 O 2 decomposition Catalysts 2021, 11, 219 9 of 17 mechanism has already been described in the literature confirming the intervention of these three species in the degradation of pollutants [78,79].  Higher reaction temperatures favor the reaction rate of ACE and the H2O2 decomposition. The decomposition of H2O2 increases significantly when the temperature increases from 25 to 55 °C, which can be explained by a higher conversion of ACE, as previously reported with other pollutants as phenol [80]. Concerning the Cr/MCM-41 catalyst, only a slight increase in the ACE removal was observed (Figure 6b) although the conversion of H2O2 was higher at higher reaction temperature (Figure 6d) due to its higher catalytic activity, with room temperature enough to achieve a total degradation of ACE. However, Fe/MCM-41 exhibited a lower catalytic activity and a higher temperature of 55 °C was necessary to reach similar results than those obtained with Cr/MCM-41. When using Fe/MCM-41 catalysts the reaction rate increases with the reaction temperature. Using the Arrhenius equation (Equation (1)) the activation energy for the acetaminophen degradation is 86 kJ·mol −1 (see Figure S7). In the case of Cr/MCM-41 the degradation is very fast at the reaction temperatures analyzed and no accurate value of the activation energy can be obtained. The Turnover Frequency (TOF) (Equation (2)) has been calculated as the mols of acetaminophen converted per mol of metal in the cata-  increases significantly when the temperature increases from 25 to 55 • C, which can be explained by a higher conversion of ACE, as previously reported with other pollutants as phenol [80]. Concerning the Cr/MCM-41 catalyst, only a slight increase in the ACE removal was observed (Figure 6b) although the conversion of H 2 O 2 was higher at higher reaction temperature (Figure 6d) due to its higher catalytic activity, with room temperature enough to achieve a total degradation of ACE. However, Fe/MCM-41 exhibited a lower catalytic activity and a higher temperature of 55 • C was necessary to reach similar results than those obtained with Cr/MCM-41. When using Fe/MCM-41 catalysts the reaction rate increases with the reaction temperature. Using the Arrhenius equation (Equation (1)) the activation energy for the acetaminophen degradation is 86 kJ·mol −1 (see Figure S7). In the case of Cr/MCM-41 the degradation is very fast at the reaction temperatures analyzed and no accurate value of the activation energy can be obtained. The Turnover Frequency (TOF) (Equation (2) The stability of the catalysts was tested in continuous tests for seven days. Figure 7 shows the evolution of ACE concentration and H 2 O 2 conversion in those long-run tests. The two catalysts are very stable under the reaction conditions for seven consecutive days. The catalytic activity of each material was stable without losing their performance. The conversion of H 2 O 2 was constant during the seven days of treatment, which results in a total degradation of ACE. To completely assess that the catalyst does not suffer significant modification after long reaction times, additional characterization of the used catalyst is necessary. However, the results shown here allow us to state that the catalyst is capable to maintain the high acetaminophen conversion at those extremely long reaction times. Metal leaching has been studied in order to confirm the possible contribution of homogeneous Fenton to the degradation of ACE. These results confirm the stability of the Cr/MCM-41 and Fe/MCM-41 catalysts prepared by a simple process which consists in a partial exchange between metal and CTA + followed by calcination. It must be replaced by the next one where "H2O2" has been replaced by "H2O2" with subscripts in Y axys of figure 7b and in the legends of figures 7a and 7b. The mechanism of the oxidation of acetaminophen has been extensively studied in the literature (Figure 8). In the case of CWPO of acetaminophen, the reaction pathway begins with the formation of benzoquinone, which result upon further oxidation in the generation of benzaldehyde and benzoic acid [81][82][83]. These by-products further decompose in smaller aliphatic organic acids such as maleic, oxalic, acetic and formic acids, which are non-toxic products [82,83]. The last step involves the complete mineralization to CO 2 , H 2 O and NO 3 − . The mechanism of the oxidation of acetaminophen has been extensively studied in the literature (Figure 8). In the case of CWPO of acetaminophen, the reaction pathway begins with the formation of benzoquinone, which result upon further oxidation in the generation of benzaldehyde and benzoic acid [81][82][83]. These by-products further decompose in smaller aliphatic organic acids such as maleic, oxalic, acetic and formic acids, which are non-toxic products [82,83]. The last step involves the complete mineralization to CO2, H2O and NO3 -.

Catalyst Preparation
Mesoporous MCM-41 support was synthesized by a hydrothermal method as reported elsewhere [29,30,84]. The molar composition used in the formulation was Si/Al = 40, 0.12 CTAB, 0.25 NaOH, 1.5 EtOH and 100 H 2 O. Briefly, an aqueous solution with CTABr surfactant and ethanol was prepared at a temperature of 35 • C. After homogenization of the mixture, TEOS was added dropwise at the same time with a solution containing distilled water, sodium aluminate and NaOH in the solution containing the surfactant. The reaction mixture was placed under constant stirring for 3 h at 35 • C, after the product was hydrothermally treated at 100 • C for two days. The final product was filtered, washed several times with distilled water and then dried at 60 • C. Different transition metals (Cu, Cr, Fe and Zn) were exchanged on the MCM surface in order to obtain the CWPO catalysts. About 2 g of as-synthesized MCM-41 was dispersed in 100 mL of 0.1 M solutions of the different metal salts and stirred for 2 h at room temperature. Then the mixture was filtered, washed several times with distilled water and dried overnight at 60 • C. For the release of the porosity of the obtained materials through the removal of the organic template and the condensation of the silanol groups, the product was calcined for 6 h at 550 • C in an air atmosphere (to improve the formation of metal oxides in air oxygen). The catalysts were named as M/MCM-41 being M the metal exchanged (Cr, Fe, Cu and Zn) [49,85].

Catalysts Characterization
XRD powder diffraction patterns of calcined samples were obtained with a Bruker AXS D-8 diffractometer using Cu-Kα radiation at 2θ = 1-60 • . Nitrogen adsorption-desorption isotherms were obtained at −196 • C using ASAP 2020 Micromeritics apparatus. The samples were previously outgassed at 100 • C for 10 h prior to the sorption measurements. The thermal properties and mass losses were followed by means of a thermobalance (TGA DISCOVERY SDT-650) under nitrogen flow. The functional groups of the metal-modified MCM-41 before and after calcination were analyzed by Fourier transform infrared (FTIR) spectra in the range of 500-4000 cm −1 using a BRUKER ALPHA Platinum-ATR instrument. The metal coordination state with the mesoporous silica was analyzed by ultraviolet visible (UV-vis) absorbance spectra using SHIMADZU UV-2700 spectrometer. The amount of metal in each sample was determined by X-ray fluorescence using a XEPOS spectrometer (Spectro Ametek). The XPS measurements were carried out on a Kratos Axis Ultra using AlKα (1486.6 eV) radiation. High-resolution spectra were acquired at 20 eV pass energy with energy resolution of 0.9 eV, the C1s line of 284.5 eV was used as a reference to correct the binding energies for charge energy shift.

Adsorption Tests
The adsorption tests were performed in batch reactor at 25, 40 and 55 • C. Briefly, 1 g of catalyst was added to a 5 mg/L of ACE solution (1000 mL) at a fixed pH close to 3 using a 1 M HCl solution. The samples were taken at different times in maximum 2 h and they were analyzed by high-performance liquid chromatography (HPCL) in order to obtain the adsorption capacities of catalysts. The results show that adsorption equilibrium was obtained after 1 h with very low adsorption capacities.

Catalytic Tests
CWPO tests were carried out in batch reactor at controlled temperature. A catalyst concentration of 1 g/L was used while the ACE initial concentration was fixed at 5 mg/L. The pH was adjusted to 3 using a 1 M HCl solution. The mixture was stirred for 1 h in order to achieve adsorption equilibrium and then the stochiometric amount of H 2 O 2 was added to the solution to initiate the reaction. The conversion of ACE was followed by HPLC (Shimadzu Prominence-I LC-2030C with a diode array detector) at a wavelength of 246 nm using a reverse phase C18 column. A mixture of acetonitrile/acetic acid 0.1% v/v was used as mobile phase (0.7 mL/min). The effects of the metal active phase, catalyst mass (0.5, 1.0 and 2.0 g/L), and reaction temperature (25, 40 and 55 • C) were analyzed.
The stability of catalysts was investigated using a continuous reaction test for one week. An ACE solution was continuously fed to the reactor (0.243 mL/min) using the best catalysts (100 mg) at a reaction temperature of 25 • C. The same volume of solution was extracted and analyzed to follow the ACE conversion with time on stream. The H 2 O 2 concentration in the reaction medium was quantified by colorimetric methods with a Cary 60 UV-Vis (Agilent Technologies) spectrophotometer, using the titanium oxisulphate method [86].

Conclusions
The mesoporous material MCM-41 with a Si/Al ratio = 40 was prepared by a hydrothermal method. The as-synthesized MCM-41 was ionic exchanged with several transition metals such as Fe, Cr, Cu and Zn and then calcined at 550 • C. The results (XRD, XPS, UV-vis, FTIR) showed that Fe, Cr, Cu and Zn were incorporated into the MCM-41 framework, which was confirmed by the inflation of the mesh parameter (a 0 of the modified materials was higher compared to the parent material), and a part of the metals were transformed in their corresponding oxides form (Fe 2 O 3 , ZnO, CuO and Cr 2 O 3 ). The nitrogen adsorption at 77K showed that the surface area was reduced due to the pores blockage by metal oxides. However, different behavior was observed for the catalyst Cu/MCM-41 in which the formation of CuO thin layers slightly decreased the surface area.
Application of these solids in the oxidation of ACE showed that the ACE conversion varies in the following sequence Cr/MCM-41 > Fe/MCM-41 > Cu/MCM-41 > Zn/MCM-41. The effect of the initial concentration of ACE, reaction temperature and the catalyst mass were studied in which the Cr/MCM-41 and Fe/MCM-41 catalysts were judged as the best catalysts for this kind of reaction. The use of both these catalysts in long-time tests showed that ACE conversion was higher and stable, which confirms the high stability of the synthesized catalysts.