*3.1. Fenton-like Reactions (H2O2/catalyst)*

The homogeneous Fenton system implies the reaction of Fe2+ with H2O2 to generate hydroxyl radicals (Equation (4)), with a high reactivity and high oxidant power, capable of oxidize organics according to Equations (1)–(3).

$$\rm Fe^{2+} + H\_2O\_2 \rightarrow Fe^{3+} + HO\bullet + HO^- \tag{4}$$

The generated HO• radicals can re-combine with Fe2+:

$$\mathrm{Fe^{2+} + HO\bullet} \rightarrow \mathrm{Fe^{3+} + HO^{-}} \tag{5}$$

The ferric ions formed may decompose the hydrogen peroxide into water and oxygen, following the Equations (6)–(10), in which ferrous ions and radicals are also generated. The reaction of H2O2 with Fe3+ is referred in literature as Fenton-like reaction [52].

$$\text{Fe}^{3+} + \text{H}\_2\text{O}\_2 \rightarrow \text{Fe-OOH}^{2+} + \text{H}^+ \tag{6}$$

$$\text{Fe-OOH}^{2+} \rightarrow \text{HO}\_2\bullet + \text{Fe}^{2+} \tag{7}$$

$$\text{Fe}^{2+} + \text{HO}\_2\bullet \rightarrow \text{Fe}^{3+} + \text{HO}\_2^{-} \tag{8}$$

$$\text{Fe}^{3+} + \text{HO}\_2\bullet \rightarrow \text{Fe}^{2+} + \text{O}\_2 + \text{H}^+ \tag{9}$$

$$\rm{HO}\bullet + \rm{H}\_{2}\rm{O}\_{2} \rightarrow \rm{H}\_{2}\rm{O} + \rm{HO}\_{2}\bullet \tag{10}$$

Other reactions involving radicals in the Fenton process are:

$$\rm H\_2O\_2 + HO\bullet \rightarrow HO\_2\bullet + H\_2O\tag{11}$$

$$\rm{HO\_2\bullet} + \rm{HO\_2\bullet} \rightarrow \rm{H\_2O\_2} + \rm{O\_2} \tag{12}$$

Notice that by reaction (11) H2O2 acts as sink for HO•, diminishing the oxidizing power of the Fenton reactants.

The homogeneous Fenton system, which implies the reaction of Fe2+/Fe3+ in solution with H2O2, has several drawbacks. By one hand, the chemical reactivity of iron is strictly dependent on the pH and only at pH ≈ 3, all three Fenton-active species of Fe2+, Fe3+ and Fe(OH)2+ coexist together. On the other hand, the final effluent contains high metal concentrations, which have to be recovered by additional treatment. In the heterogeneous systems Fe2+ and Fe3+ are part of a solid, which results in different advantages, especially related to recovery of the catalyst and the low leaching of ions.

The Fenton-like reactions involved in a heterogeneous system are the following:

$$\rm{S-Fe^{2+} + H\_2O\_2 \to S-Fe^{3+} + HO\bullet + HO^-} \tag{13}$$

$$\text{S-Fe}^{3+} + \text{H}\_2\text{O}\_2 \rightarrow \text{S-Fe}^{2+} + \text{HO}\_2\text{O} + \text{H}^+ \tag{14}$$

$$\text{S-Fe}^{2+} + \text{HO}\bullet \rightarrow \text{S-Fe}^{3+} + \text{HO}^- \tag{15}$$

$$\text{S-Fe}^{3+} + \text{HO}\_2\bullet \rightarrow \text{S-Fe}^{2+} + \text{O}\_2 + \text{H}^+ \tag{16}$$

$$\text{Organic} + \text{HO} \bullet \rightarrow \text{R} \bullet + \text{H}\_2\text{O} \rightarrow \dots \rightarrow \text{CO}\_2 + \text{H}\_2\text{O} \tag{17}$$

where S represents the surface of the catalyst. The reaction (14) is rate-limiting since its rate constant is ca. four orders of magnitude lower than that of reaction (13).

Although the classical Fenton system is based on the use of Fe2+/H2O2, other elements with multiple redox sates (like chromium, cerium, copper, cobalt, manganese and ruthenium) can directly decompose H2O2 into HO• through conventional Fenton-like pathways [53]. Therefore, perovskites containing these elements, mainly in B position, can be used in this kind of reaction.

Rhodamine B (RhB) is one of the most studied organic pollutants in water in the Fenton-like reactions [54–58]. The first study reporting the use of perovskites as catalysts for the removal of RhB in Fenton-like reactions was carried out by Luo et al. [54]. In this work, BiFeO3 magnetic nanoparticles (BFO MNPS) prepared by sol-gel method were tested in the degradation of RhB in the presence of H2O2 at 25 ◦C and pH = 5. According to the isoelectric point of BFO MNPs (I.P. = 6.7), under these conditions, the anionic form of the dye (pKa = 3.7) interacts easily via electrostatic forces with the positively charged catalyst particles. By selecting initial H2O2 and catalyst concentrations as 10 mM and 0.5 g/L, respectively, 95.2% of RhB was degraded in 90 min and a TOC removal of 90% was achieved within 2 h, in contrast to the removal of only 10% of RhB in the presence of Fe3O4 nanoparticles. By Montecarlo (MC) simulations authors concluded that after the adsorption of H2O2 molecules on the surface hollow sites of BFO MNPs facets, they are activated to generate HO• radicals, which then decompose RhB into other smaller organic compounds and CO2. BFO MNPs showed excellent chemical stability during reaction (as checked by XPS), being reusable for at least five cycles, without a significant loss of activity. BFO MNPs were also tested in the degradation of methylene blue and phenol, leading to 79.5% and 82.1% of removal, respectively.

Zhang et al. [55] synthesized a series of Cu-doped LaTiO3 perovskite (LaTi1−xCuxO3, x = 0.0–1.0) by a sol-gel method, which resulted be very efficient for the degradation of RhB with H2O2 in a pH range of 4–9. In contrast to the absence of activity of sample containing only titanium, the coexistence of Ti3+/Ti4+ and Cu+/Cu2+ in the perovskite structure of partially substituted samples allowed the degradation of 8 mg/L of RhB through redox cycles involving the transformation of H2O2 into HO• and HO2•/O2•−. For a H2O2 concentration of 10 mM, about 84% of RhB was decolorized within 2 h in the presence of 1.4 g/L of LaTi0.4Cu0.6O3. Notice that the amount of RhB degraded was slightly lower than the observed by Luo et al. [54]; although the catalyst amount used in Reference [55] was almost three times higher, the initial concentration of RhB was approximately the double. The reduction of H2O2 to O2, which is carried out by oxygen vacancy [22], was not observed in this reaction.

The surface area of perovskites is low and as a consequence, the interaction between the contaminants and the active sites is limited. In order to improve the catalytic efficiency of perovskite-like oxides by increasing their surface area, some strategies have been developed, such as their supporting on mesoporous silica supports [56–58] or in honeycombs [59] and the formation of nanocomposites [60,61].

In this sense, La-FeO3/SBA-15 [56] was more efficient than non-supported LaFeO3 for catalysing RhB oxidation in the presence of H2O2 under ambient conditions due to a synergic effect between the large capacity of mesoporous SBA-15 for RhB adsorption and the high number of active sites exposed in LaFeO3 nanoparticles for reacting with H2O2. The best catalyst was the sample containing many oxygen vacancies (as deduced from XPS results), which are a key factor influencing the performance of these catalysts in oxidation reactions. The catalyst was efficient in a wide pH range (2–10). Under the optimum conditions, a degradation of RhB of 87% was achieved after 3 h. No leaching of Fe3+ was observed in the solution after reaction, the contribution of homogeneous Fenton reaction being discarded. The stability of La-FeO3/SBA-15 was also confirmed by carrying out four cycles of reutilization, which showed no deactivation of the sample. The catalyst was also applied for the degradation of other organic dyes, achieving a decomposition of 66% for methylene blue and 42% for brilliant red X-3B and direct scarlet 4BS.

The good synergy between the support and the LaFeO3 perovskite was explored by the same authors [58], who tested different supports based on mesoporous silica, such as SBA-15, SBA-16 and MCF and on nanosized silica powders (NSP). Different factors influence on the catalytic behaviour for degradation of RhB. By one hand, the RhB adsorption on the support is a crucial step of the reaction and as a result, the combination of LaFeO3 with a non-porous support showing a low capacity of adsorption decomposed the RhB in a little extent. On the other hand, a network of pores with short length is necessary to allow the transportation of RhB to the active sites of LaFeO3. In this sense, the shorter the pore length, the faster the RhB molecules reached the catalytic centres and were oxidized (see the transport process in Figure 2). Authors concluded that LaFeO3 supported on MCF containing randomly distributed pores with short length was the best catalyst for oxidative degradation of RhB in aqueous solution, achieving a removal of the contaminant of 97% in 2 h.

SBA-15 was also used by the same authors as support of a perovskite-type oxide La2CuO4 containing a few amounts of CuO [57]. The solid was tested in the degradation of RhB and organic dyes, including reactive brilliant red X-3B, direct scarlet 4BS and methylene blue under ambient conditions. The catalyst was active in a wide pH range (2–10) and depletion of RhB between 85% and 95% was produced after 3 h, depending on the amount of catalyst. The mineralization of RhB into CO2 was completed and the catalyst could be recycled. Although the activity decreased in ca. 14% in the fifth cycle, it could be recovered after a treatment of the used catalyst in air at 500 ◦C for 2 h.

**Figure 2.** A proposed scheme of transporting RhB from the solution to the pore and then to the surface-active site over LaFeO3 catalysts supported on porous SBA-15, SBA-16 and MCF. With permission from [58].

Another approach for modifying the surface properties and reactivity of perovskites is the formation of nanocomposites [60,61]. In this regard, a novel 3D perovskite-based composite BiFeO3/carbon aerogel (BFO/CA) prepared by sol-gel method led to a 95% of degradation of ketoprofen in 150 min and a TOC removal of 60% after 5 h [60]. These activities values were significantly higher than those obtained for bulk BFO and nano BFO, due to the higher reducibility of Fe3+ and Co3+ species in the composite, as deduced from TPR studies and to the dispersion of active sites not only on the surface of CA support but along the 3D structure of CA. Furthermore, the catalyst was active in a wide pH range of 3–7 and the leaching of iron was low.

A La1+xFeO3 (L1+xFO, 0 ≤ x ≤ 0.2) nanocomposite formed between LaFeO3 and an inert La2O3, resulted to be twice more active for degradation of methyl orange that the pristine LaFeO3 [61]. The modification of surface properties, such as surface Fe2+ concentration, surface defects, H2O2 adsorption capacity and charge-transfer rate led to an enhanced Fenton-like activity in the composite. The most notorious aspect of this work was that the major reactive species were not hydroxyl radicals but singled oxygen (1O2), as deduced from in situ electron paramagnetic resonance analysis and radical scavenging experiments. Authors proposed the corresponding mechanism of 1O2-based composite/H2O2 system. 100% of contaminant was degraded in 90 min at pH = 3 and a total organic carbon (TOC) removal of 96% was achieved after 4 h.

Other contaminants degraded by LaFeO3 perovskite, in this case auto supported, were different pharmaceutical and herbicides [62]. Among them, sulfamethoxazole (SMX) was completely removed in LaFeO3–H2O2 system after 2 h at neutral pH. By formation of a surface complex between LaFeO3 and H2O2, the O-O bond in H2O2 is weakened and chemical environment of iron changes, the Fe3+/Fe2+ redox potential decreasing significantly, which accelerates the cycle of Fe3+/Fe2+ and produces more HO• and O2•−/ HO2• radicals, enhancing the Fenton-like removal of organic compounds. The TOC removal was 22% in 2 h and SMX was transformed into simpler aliphatic acids, mostly biodegradable.

Due to its abundance in most of wastewater effluents and its toxicity, phenol is a usual organic compound model in developing methods for water remediation, including AOPs. The removal of phenol and phenolic compounds has been tested in Fenton-like reactions on different perovskites, mainly containing iron or copper in B position [63–65]. LaFeO3 and BiFeO3 were tested by Rusevova et al. [63] in the degradation of phenol. The influence of reaction temperature, catalyst and H2O2 concentrations and pH, on the catalytic behaviour was studied. The rate constant for phenol degradation, which increased with temperature, was 3-fold higher when initial reaction pH diminished from 7 to 5. Conversion values of phenol of 90–95% were achieved after 6 h and leaching of metals was negligible. The most new-fangled aspect of the study was that in order to settle the nature of the active oxidizing species authors used compound specific stable isotopic analysis (CSIA) as alternative to other conventional techniques. Based on their results, authors concluded that the major species involved in

phenol degradation were hydroxyl radicals. They extended the application of both LaFeO3 and BiFeO3 to the removal of methyl terc-butyl ether (MTBE), for which a depletion of 80% was obtained after 6 h.

Different perovskite-like oxides LaBO3 (B = Cu, Fe, Mn, Co, Ni) synthesized using the Peccini method were tested in Fenton-like degradation of phenol but only LaCuO3 and LaFeO3 were active [64]. Authors studied the recyclability of the catalysts during 3 cycles for LaCuO3 and 40 cycles for LaFeO3.The induction period observed in the first cycle for LaFeO3 was significantly shortened for the second and successive cycles. In this sense, a degradation of phenol of 75% was produced in 5 h in the second cycle, in contrast to the 22% observed in the first one. The reasons for this improvement in the activity were an increase in the surface concentration of oxygen containing species (water and carbonate) involved in the transformation and the formation of dispersed particles of iron oxides on the surface. The TOC conversion of 21–22% after 10 h did not change for the different cycles.

Hammouda et al. [65] prepared ceria perovskite composites CeO2-LaCuO3 and CeO2-LaFeO3, which were more active for the degradation of bisphenol than non-doped perovskites, especially at short reaction times. Furthermore, CeO2 improved the stability of perovskites towards leaching of metals. Authors attributed the enhancement in the activity to the fact that, as observed by XPS, more Ce3+ ions were formed in the ceria-perovskite catalysts, due to an electron transfer from the transition metal of perovskites to the CeO2. As a result, more oxygen radicals were formed by interaction of H2O2 with Ce3+, favouring the Fenton-like degradation of the contaminant. By following the evolution of the intermediates formed, authors proposed a mechanism of reaction and a degradation pathway.

In order to improve the catalytic ability of BiFeO3 nanoparticles to degrade recalcitrant pollutants, some authors have proposed an in-situ surface modification by using chelating agents [66,67]. In this regard, the bisphenol A (BPA) degradation in a wide pH range (4–9) was accelerated when the nano-BiFeO3 were modified by adding different ligands to the Fenton solution, such as tartaric acid, formic acid, glycine, nitrilotriacetic acid and ethylenediaminetetraacetic acid (EDTA) [66]. EDTA was the most efficient chelating agent, mainly because of a higher HO• formation from the H2O2 decomposition. Under the optimum conditions 91.2% of BPA was removed within 2 h, in contrast to the 20% of BPA degraded with unmodified BFO. Although the use of chelating agents increased the contribution of Fenton homogeneous reaction by formation of soluble iron complexes, the trend observed in the BPA degradation for reactions carried out with different ligands did not follow the order of leached ions, indicating the irrelevant contribution of homogenous reaction to BPA degradation. As EDTA was the most efficient chelating agent, it was also used by same authors [67] as ligand for BiFeO3 in the degradation of triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol), a broad-spectrum antibacterial agent widely used in personal and health care products. When pristine BFO were used, triclosan was mainly transformed into 2,4-dichlorophenol, a carcinogenic compound. The addition of EDTA modified significantly the dechlorination ratio of triclosan, which increased from 26.4% for H2O2-BFO sample up to 97.5% in the chelated system. Triclosan was degraded almost completely in 3 h under the optimal conditions.

Another strategy to improve catalytic activity of perovskites in AOPs is the hetero-doping in order to produce more active sites of the low-valence B-site transition metals (i.e., Fe2+, Cu+ and Ti3+) or to introduce oxygen vacancies, which can facilitate the transformation of H2O2 into HO• [39,40]. In this sense, some perovskites containing partially substituted manganese in B position, have been tested in Fenton-like reactions for the degradation of methylene blue (MB) [68], different dyes [69] and paracetamol [70].

Maghalaes et al. [68] tested LaMn1-xFexO3 and LaMn0.1-xFe0.9MoxO3 perovskites in the decomposition of H2O2 to O2 and in the oxidation of MB. The presence of manganese in the perovskites seemed to play an important role on the H2O2 decomposition rate, which decreased with the amount of Mn substituted by Fe and/or Mo. However, LaMnO3 was not active for the MB discoloration, which suggested that it was able to transform the H2O2 into O2 but it was unable to form the HO• radicals, necessary to degrade the dye molecules. On the contrary, samples substituted by Mo degraded MB up to 20% in 1 h.

Jahuar et al. [69] synthesized a series of manganese-substituted lanthanum ferrites having compositions LaMnxFe1-xO3 (x = 0.1–0.5) by a sol–gel auto-combustion method, which were used as catalysts in the removal of anionic dyes (Remazol Turquoise Blue, Remazol Brilliant Yellow) and cationic dyes (MB, Safranine-O) by the action of H2O2, in the absence and presence of visible-light. The initial pH of solution was fixed in all cases to the value of 2. Unsubstituted LaFeO3 produced a low dye degradation for long time periods, exhibiting a poor catalytic activity under dark conditions. However, the partial substitution of iron by manganese led to catalysts able to degrade over 90% of dye in time periods of 150–300 min, due to the Fenton-like activity of manganese ions, capable of existing in various oxidation states. In the presence of light, an enhancement in the catalytic activity was produced and degradation times were reduced to 25–70 min.

The contribution of manganese ions to Fenton-like reaction was, on the contrary, discarded by Carrasco-Díaz et al. [70] in the decomposition of paracetamol by H2O2 under mild reaction conditions (25 ◦C and pH ≈ 6) in the presence of LaCuxM1-xO3 (0.0 ≤ x ≤ 0.8, M = Mn, Ti) perovskite-like oxides prepared by amorphous citrate decomposition. Degradation values of paracetamol between 80% and 97% were achieved after 5 h. XPS studies of the catalysts allowed authors to conclude that Cu2+/Cu+ were the catalytically active species, the catalysts containing a higher amount of copper at the surface, mainly as Cu2+, being the most active. The titanium and manganese species seemed not to be responsible of the enhanced activity observed in some of the substituted samples with respect to that of LaCuO3. The catalysts were recyclable for at least three cycles and a negligible leaching of metals was produced. TOC values of 47–54% were achieved.

Finally, some mathematical analysis of the heterogeneous oxidations of contaminants by perovskites have been carried out. More concretely, mathematical modelling of photo-Fenton-like oxidation of acetic acid by LaFeO3 has been reported [71,72]. From the experimental results authors concluded that the main reactions occurring in the system were the complete mineralization of acetic acid by H2O2 due to the presence of the catalyst and the decomposition of H2O2 into water and O2 in the homogeneous phase. Therefore, this kind of reaction should not be considered as an AOP, because no hydroxyl radicals were formed.

Table 1 summarizes the conversion values and reaction conditions for the use of perovskites in the degradation of different organics by Fenton-like reactions and photo Fenton-like reactions, these last being revised in the following section.




