*3.2. Photo Fenton-Like Reactions (H2O2/Catalyst/Light)*

In the photocatalytic oxidation processes, the electron–hole pairs in the catalyst are produced via the irradiation of the UV light and the oxidative radicals are formed between the catalyst and water interface [82]. The formation rate of HO• radicals in photo-Fenton processes is higher than in Fenton processes. While the Fenton reaction is governed principally by Equation (4) leading to the formation of HO• radicals, in the Photo-Fenton process occurs, in addition, the photolysis of H2O2:

$$\text{H}\_2\text{O}\_2 + \text{hv} \rightarrow 2\text{ HO}\bullet\tag{18}$$

and the photo reduction of Fe3+:

$$\text{Fe}^{3+} + \text{H}\_2\text{O} + \text{hv} \rightarrow \text{HO}\bullet + \text{Fe}^{2+} + \text{H}^+ \tag{19}$$

Different perovskite oxides, non-supported [69,73–76], supported on monoliths [59,77–80] or in form of composites [81], have been used as catalysts for the degradation of several organics in the presence of H2O2 under light irradiation conditions.

A LaMnO3 perovskite prepared by co-precipitation method resulted to be an excellent photo-Fenton catalyst of oxidation of phenol [73]. The phenol conversion (99.92% in 4 h) obtained when LaMnO3 was activated by UV radiation in the presence of stoichiometric amount of H2O2 necessary to degrade phenol, was even higher than the achieved when using TiO2 as catalyst (98% in 4 h). Furthermore, LaMnO3 could be regenerated by calcination after reaction, yielding to similar catalytic performance to that of the first cycle.

In Reference [75] a series of iron perovskites containing europium in B position, EuFeO3 (EFO), calcined at different temperatures, was used for the photodegradation of RhB by combination of visible light and H2O2. By the action of visible light, electron-hole pairs were formed in EFO nanoparticles and electrons were easily trapped by H2O2, leading to the formation of HO• radicals. In addition, a complex between Fe3+ at the surface (≡Fe3+) and H2O2 was formed and ≡Fe3+ was transformed into ≡Fe2+, generating HO• and HO2•, which decomposed RhB. Authors studied the effect of the calcination temperature of the catalysts on the band gap, microstructure and photocatalytic activity. The perovskite calcined at 750 ◦C showed the best catalytic behaviour, due to the combination of good crystallinity and appropriate BET surface area and band gap. The photodegradation of 37% of RhB after 3 h increased up to the 71% when H2O2 was added, which proves the Fenton-like activity of EFO nanoparticles.

One strategy to improve photocatalytic activity of ABO3 perovskites is the substitution of the element in A or B position, which leads to the introduction of defects into the narrow band gap and to the formation of oxygen vacancies, which inhibit the recombination between the photogenerated electrons and the holes. In this regard, 2-chlorophenol (2-CP), was degraded in the presence of three metal doped BiFeO3 (BFO) nanoparticles, H2O2 and visible light [74]. BiFeO3 perovskite was substituted either in A position (Bi0.97Ba0.03FeO3) or B position (BiFe0.9Cu0.1O3) and in both (Bi0.97Ba0.03 Fe0.9Cu0.1O3). After only 70 min of visible light irradiation, Cu-doped BFO and Ba-Cu co-doped BFO almost completely removed 2-CP. The mineralization degree reached was of 68% and 73%, respectively. Authors concluded that in addition to the participation of Fe2+/Cu+ couple active for the formation of HO•, the oxygen vacancies on the surface can also participate by activating H2O2 molecules to form a lattice oxygen, which is furtherly desorbed as O2.

More recently Phan et al. [76] studied the efficiency of LaFeO3 (LFO) perovskites, doped with Cu in B position (LaFe1-xCuxO3), in the photo-Fenton decolorization of methyl orange (MO). Interestingly, the substitution of Cu into Fe-site in LFO modified the light absorption property of perovskite, as noted in the UV–vis absorption spectra and the corresponding band gap energy of LFO and LaFe1-xCuxO3 shown in Figure 3a,b, respectively. Notice that all the samples had suitable band gap energy for organic pollutant degradation under visible light irradiation. When Fe was substituted by 15 mol% of Cu

(LFO-15Cu), the MO degradation rate was improved in ca. 60% and 92.9% of MO was removed in only 1 h at an initial solution pH of 6. Under the optimum conditions, the photocatalytic performance of LFO-15Cu was also evaluated for two cationic dyes, rhodamine B (RhB) and methylene Blue (MB), obtaining even better results: 99.4% of degradation for RhB and 98.8% for MB in 60 min.

**Figure 3.** (**a**) UV–vis absorption spectra and (**b**) corresponding band gaps of LFO and LaFe1-xCuxO3. With permission from [76].

LaFeO3 or Pt/LaFeO3 perovskites supported on honeycomb monoliths have been tested in the degradation of tartrazine, a not-biodegradable dye used in food industries [59] by continuous flow of H2O2 in the presence of UV light at different pH values. The natural pH of solution (near 6) was the best operating condition, under which 100% of tartrazine was discoloured after 30 min of irradiation and mineralization was complete after 40 min. On the contrary, the discoloration was 50% under acidic condition (pH 3) and 55% under basic condition (pH 9), after 30 min of irradiation. In the first case, the excess of H+ could react with HO• and produce water subtracting hydroxyl radicals necessary for the decomposition of tartrazine [83]. In the second one, H2O2 was accumulated in liquid medium because under alkaline conditions H2O2 has a very high stability [84] and as a consequence, the production of hydroxyls radicals was limited.

Supported LaMeO3 (Me= Mn, Co, Fe, Ni, Cu) perovskites were prepared by impregnation of thin wall of monolithic honeycomb cordierite support with different active phase loadings and tested in the photo-Fenton oxidation of acetic acid [77]. In the case of LaMnO3 sample, the honeycomb was also impregnated with 0.1% of Pt. Photo-Fenton activity was closely related to the amount of active phase supported on monolithic carrier and LaFeO3 and Pt/LaMnO3 perovskites were the best catalysts in terms of reaction rate. The addition of Pt enhanced the initial rate of acetic acid degradation, achieving the highest TOC removal, 18% in 1 h; however it did not enhance the catalytic performance after 5 h.

Different loads of LaFeO3 perovskites supported over corundum monoliths were studied by the same authors in the photo-Fenton degradation of several organics [78] under UV irradiation. 97% of TOC removal was attained in the degradation of acetic acid after 4 h with the catalyst containing 10.64 wt% of LaFeO3 and values of 53, 62 and 95% were achieved when ethanol, acetaldehyde and oxalic acid, respectively, were used as model pollutant.

The excellent catalytic behaviour of this catalyst was extended to other contaminants, as methyl terc-butyl ether (MTBE) [79]. About 100% of TOC removal and complete mineralization of MTBE into CO2 and water was achieved if H2O2 was continuously dosed during irradiation time of 2 h at solution pH of 6.7. Although the TOC removal obtained by the combination of H2O2 and UV light in the absence of catalyst was quite high (about 97%), a very significant formation of CO was observed, indicating the importance of LaFeO3 for the improvement in the mineralization of MTBE.

The degradation of methylparaben was studied in the presence of four ABO3 perovskite catalysts (A: La, Bi and B: Fe, Ti-Fe) supported on a monolithic structure [80]. BiFeO3 was the best catalyst and under the optimum conditions 82.8% of pollutant was degraded in only 90 min. In the absence of UV light, only 10% of methylparaben was removed. An interesting aspect of this work was the study of toxicity carried out by cress seed, showing an inhibition of only 1.09% in the growing of roots, which demonstrated the low toxicity of the products of degradation.

The combination of photocatalytic activity and oxidizing power of H2O2 was also applied for the degradation of tetrabromobisphenol A (TBBPA) with a graphene-BiFeO3 composite as catalyst [81]. The catalytic activity was influenced by calcination temperature, pH, presence or not of EDTA, dosage of H2O2 and load of catalyst. The degradation of TBBPA approximately followed a kinetics of pseudo first order. Under the most favourable conditions and in the presence of EDTA the rate constant of TBBPA degradation with the graphene-BiFeO3 was 5.43 times higher than that of BiFeO3 and 80% of TBBPA was removed in 15 min. This enhancement in the catalytic activity was attributed to the increasing of the adsorption capacity (due to a large surface area) and to the high electron transfer ability of graphene in the composite, which favoured the generation of reactive species. The composite was stable and could be reused for five cycles without loss of catalytic activity.
