**1. Introduction**

Advanced oxidation processes (AOPs) are based on the generation of radical intermediates, mainly hydroxyl radicals (HO•), in amount enough to be able to attack and oxidize either partially or fully most of the recalcitrant chemicals present in the effluent water, such as pesticides, dyes, pharmaceuticals and so on [1]. The processes based on the free radicals occur at higher rates of degradation than those based on other chemical oxidation technologies and are not highly selective [2,3]. The high oxidation potential of hydroxyl radicals (2.80 v) make them capable of attacking organic compounds by abstracting a hydrogen atom or by adding to the double bonds, carrying out their mineralization by transformation into more oxidized intermediates, carbon dioxide, water and inorganic salts. These reactions of hydroxyl radicals with organic compounds can be written as follows:

$$\text{H}\text{HO}\bullet + \text{R-H} \rightarrow \text{H}\_2\text{O} + \text{R}\bullet \tag{1}$$

$$\text{HO}\bullet + \text{C=C} \rightarrow \text{HO-C-C} \bullet \tag{2}$$

$$\text{H}\text{HO}\bullet + \text{Ph-H} \rightarrow \text{Ph-H}(\text{OH})\bullet \tag{3}$$

AOPs can be classified in several categories, depending on the different reagent systems used for the generation of hydroxyl radicals. Attending to the reaction medium, these advanced oxidation processes can be classified either as homogeneous or heterogeneous [4]. The first ones can be subdivided in turn in those using energy (ultraviolet or visible radiation, ultrasound energy, electrical energy) and those not involving energy (ozone (O3) in alkaline medium, O3/H2O2 and H2O2/homogeneous catalyst, generally Fe2+, known as Fenton process). The heterogeneous processes can be classified in four main groups: (i) catalytic ozonation, which uses the combination of O3 and a

solid catalyst; (ii) photocatalytic ozonation, under the action of O3/light (UV or visible)/solid catalyst; (iii) Fenton-like processes, which are produced by the action of H2O2/solid catalyst, containing mainly the Fe2+/Fe3+ couple but also other transition metal ions with multiple oxidation states; when they are combined with the action of light they are called Photo-Fenton processes; and iv) photocatalytic oxidation, by combination of light (UV or visible) and a solid catalyst.

It must be remarked that the single ozonation or the use of only H2O2 belong to the class of chemical oxidation technologies, as they work on the direct attack of the oxidants, not being considered as AOPS, because they do no generate hydroxyl radicals by themselves [5]. Only when O3 and H2O2 are combined between them or its individual action is supplemented by other dissipating energy components, such as UV/visible light or ultrasound or by activation with a catalyst, the formation of free radicals occurs and they can be considered AOPs.

In the last fifteen years some reviews in literature have devoted to show the state of art of AOPS for wastewater treatment [4–7] and in particular Fenton and photo-Fenton processes [8,9]. Most of these oxidation technologies are usually expensive and in addition, they are unable to completely degrade the organic compounds present in real wastewater and they cannot process the large volumes of waste generated. However, AOPs can degrade the residue up to a certain level of toxicity and then the intermediate can be furtherly degraded by the conventional methods. Furthermore, the combination of AOPs (as a pre-treatment or post-treatment stage) with a biological treatment contributes to reduce operating costs of the global process [10].

As mentioned above, the heterogeneous advanced oxidation processes generally use solid catalysts in combination with other systems (O3, H2O2, light) to carry out the degradation of organics. The main advantage of heterogeneous catalysts with respect to the homogeneous ones is the facility of separation of the product and of the recovery of the catalyst. However, to be applied in the industry, heterogeneous catalysts must satisfy some specifications, such as high activity, thermal, mechanical, physical and chemical stability and resistance to the deactivation.

Perovskite-type oxides of the general formula ABO3, where A is a rare earth metal and B a transition metal, have attracted the attention of many scientists because of their unique structural features. They have a well-defined structure, which allows the introduction of a wide variety of metal ions in both A and B positions [11,12]. Its structure is represented in Figure 1. The partial substitution of these cations by other foreign leads to changes in the oxidation states of metal ions and to the formation of oxygen vacancies. The thermal and hydrothermal stability of perovskites is quite high and as a result, they can be applied to gas or solid reactions carried out at high temperatures or liquid-phase reactions occurring at low temperatures [13,14]. The high mobility of network oxygen and the stabilization of unusual oxidation states confer them a diversity or properties, which allows their application as solid oxide fuel cells [15], magnetic and electrode materials [16], chemical sensors [17], adsorbents [18] and heterogeneous catalysts in industrial reactions [11,14,19]. One of the potential applications of perovskites is as catalysts in carbon-based electrodes, which has incited many scientists to study the mechanism of the catalytic decomposition of H2O2 by perovskites [20–22].

Catalysts used in oxidation technologies can be classified as follows: (i) metal catalysts, usually supported on a metal oxide surface (TiO2, Al2O3, ZrO2 and CeO2) or on active carbon; (ii) metal oxide catalysts and (iii) organometallic catalysts. Different heterogeneous catalysts have been applied in some AOPs. As an example, the use of several heterogeneous systems containing iron species stabilized in a host matrix, such as oxides [23–26], clays [23,27–30], zeolites [29,31,32] or carbon materials [33–37] in the Fenton and the Fenton-like processes has been reported. In the last years perovskites have been applied as Fenton catalysts because of their versatile composition and high stability [19]. Furthermore, the existence of redox active sites in B cations and oxygen vacancies may facilitate the transformation of H2O2 into HO• [38–40].

There are some reviews in literature describing the employment of clays and mineral oxides, mainly of iron, in Fenton-like processes [23,25,26]. However, until our knowledge, there is no any review reporting the use of perovskites in AOPs. In the present paper we describe the utilization of

perovskites and like-perovskite oxides in this kind of processes, classified as a function of the oxidant reactant responsible for generating the free radicals, by alone or combined with other systems, giving place to the hybrid methods. Thus, in first place, in Section 2, we report the processes based on ozone, O3, including catalytic ozonation (O3/catalyst) and photocatalytic ozonation (O3/catalyst/light). In the following section we revise the use of H2O2 as oxidant in three kinds or systems: (i) Fenton-like reactions (H2O2/catalyst); (ii) photo Fenton-like reactions (H2O2/catalyst/light); and (iii) catalytic wet peroxide oxidation, CWPO, (H2O2/catalyst/air). In the last fifteen years a new chemical oxidant, peroxymonosulfate (PMS), has aroused a great interest as alternative to others as H2O2 or O3. For that reason, the fourth section is devoted to describing the use of PMS, activated by a catalyst (perovskite) or by the combination of both a catalyst and light irradiation. The photocatalytic oxidation, consisting in the activation of perovskite by irradiation with UV or visible light, is revised in Section 5. Finally, a short section devoted to the degradation under dark ambient conditions, show some examples in which perovskites catalyse the oxidation of organics in the absence of light and without additional chemical oxidant. Although these processes should not be considered strictly AOPS, because radicals are not formed, some authors consider them as novel advanced oxidation technologies for low cost treatment of wastewaters.

**Figure 1.** Unit cell of perovskite centred on A.

#### **2. Processes Based on Ozone**

Heterogeneous catalytic ozonation is getting an enormous attention in the treatment of drinking and waste water, because of its capacity to improve the mineralization and degradation of organic pollutants, its manufacturing simplicity and economic nature. There are many advantages to the use of ozone compared to other conventional technologies due to its high oxidizing power (2.07 V). Ozone can degrade organic pollutants by direct electrophilic attack with molecular ozone or by indirect attack with hydroxyl radicals HO•, which are generated through its decomposition process. Single ozonation has been widely used in water and wastewater because is an effective oxidation process. However, several disadvantages can limit its application such as slow and incomplete oxidation or mineralization (the intermediates are not totally oxidised to CO2 and water) and the low solubility and stability of ozone in water. At first sight, heterogeneous catalytic ozonation does not present these drawbacks although much research is still needed. Catalytic ozonation utilizes solid catalysts in order to improve the decomposition of ozone and to enhance the production of hydroxyl radicals, HO•. In general, the heterogeneous catalysts decompose the ozone into caged or free radicals or simply they adsorb reactants facilitating their reaction. To date, several types of heterogeneous ozonation catalysts such as metal oxides, supported metal oxides and carbon materials have been tested with promising results. Perovskites-type metal oxides constitute undoubtedly an interesting alternative due to the high stability under aggressive conditions, high degree of stabilization of transition metals in their oxidation states and high oxygen mobility. However, the catalytic ozonation mechanism with perovskites is still a challenge for chemists and not many detailed studies are available. Most of the studies have been carried out by the same research group [41–44].

The first studies concerning catalytic ozonation with active perovskites appeared in 2006 [41]. The authors studied the ozonation decomposition of pyruvic acid, a refractory substance typically produced after oxidation of phenol-like compounds, in the presence of LaTi0.15Cu0.85O3 perovskite. This perovskite was active and stable in the ozonation process being oxalic and acetic acids the only intermediates formed. Experimental results clearly indicated that typical operating parameters like ozone concentration, mass of catalyst or temperature, performed a key role on the pyruvic acid ozonation. The catalyst exhibited high stability and its catalytic activity improved after the first use. For instance, after 150 min of reaction the pyruvic acid had practically been eliminated, in contrast to the 67% of conversion achieved with fresh catalyst for the same period. Regarding kinetic considerations, authors proposed a Langmuir–Hinshelwood mechanism derived from bi-adsorption of pyruvic acid and ozone on different active sites and successive reaction of pyruvic acid with the O• radicals on the surface through reactions 1 and 2, before occurring the desorption of formed products (see Scheme 1).

$$
\begin{array}{ccc}
\downarrow & & & \downarrow \\
\downarrow & & & \downarrow \\
\downarrow & & & \downarrow \\
\downarrow & & & \downarrow \\
\end{array}
$$

$$\underbrace{\begin{array}{c} \underbrace{\bigotimes\bigvee\bigvee\cdots\bigotimes\bigvee\bigvee\cdots\bigotimes\bigvee\cdots}\_{\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\cdots\bigotimes\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf A}^{\star}\bigvee\operatorname{\bf$$

**Scheme 1.** Mechanism of catalytic ozonation of piruvic acid in the presence of LaTi0.15Cu0.85O3. With permission from [41].

Carbajo et al. [42] extended the study of the removal of pyruvic acid from water through catalytic ozonation to other perovskites, LaTi1-xCuxO3 and LaTi1-xCoxO3, which were compared with other catalysts, such as Ru-Al2O3, Ru-CeO2, FeO(OH) and MCM-41, this last impregnated with copper or cobalt. The results showed that only perovskites and Ru-CeO2 catalysts increased significantly the pyruvic acid depletion with respect to the produced in the absence of catalyst. Conversion values of 80% were reached after 2 h of catalytic ozonation. Due to the low adsorption capacity of the catalysts to adsorb the pyruvic molecules, authors concluded that the catalytic ozonation mechanism was governed by surface reactions involving adsorbed ozone and dissolved pyruvic acid.

Another pioneering work with LaTi0.15Cu0.85O3, the same catalyst used in Reference [41], was carried out to eliminate gallic acid, a primary intermediate of benzoic acid oxidation [43]. The role of different operating variables was studied. Whereas the catalyst and ozone doses exerted a positive influence in the ozonation rate, the increment in the initial acid gallic concentration diminished the conversion. The activity of the catalyst in terms of acid elimination was kept for consecutive cycles. However, the catalyst displayed a partial deactivation in terms of total organic carbon (TOC) elimination after the second reuse. In any case, the TOC degree was still higher than the one achieved in the non-catalytic system.

Finally, Carbajo et al. [44] went a step further in analysing the activity of the same catalyst, LaTi0.15Cu0.85O3, in the ozonation of four real phenolic wastewaters coming from agro-industrial field, a wine distillery industry, olive debittering and from olive oil production. The main goal was to study the activity and stability of the catalyst together with the influence of the different operating variables. The results suggested that if enough time was allowed the catalytic ozonation of the phenolic mixture achieved 100% of mineralization. Moreover, the increment of temperature promoted the mineralization level.

Some of the compounds that cannot be easily removed from drinking water or wastewater by classical treatments are pharmaceutical compounds. Catalytic ozonation allows high removal of organic carbon of these compounds being the most appropriate process. As a first approach Beltran et al. [45] tested two copper and cobalt perovskites, LaTi0.15Cu0.85O3 and LaTi0.15Co0.85O3, as catalysts to remove in the presence of ozone sulfamethoxazole, a synthetic antibiotic usually found in municipal wastewaters. Some experiments were also carried out in the presence of activated carbon, as promoter of the activation of ozone. The results showed that catalytic or promoted ozonation were not necessary to eliminate sulfamethoxazole from water, because it can be removed only by the action of ozone. However, from a practical point of view, the combined ozone processes are clearly recommended in order to remove the resulting total organic carbon (TOC). The catalytic ozonation of two pharmaceutical compounds, the drug diclofenac and the synthetic hormone 17-ethynylstradiol, was also conducted on the same perovskites by the same authors [46], obtaining similar results to those observed in their previous work. Both compounds can be eliminated by direct ozonation; however, when copper perovskite was used, the TOC removal reached the 90% after 2 h of reaction.

To date, the main advantage of catalytic ozonation is the ability to improve the mineralization degree achieved at the end of the process. In this sense, non-substituted perovskites type LaBO3 (B=Fe, Ni, Co and Mn) and substituted perovskites type LaBxCu1-xO3 (B=Fe and Al), have been proposed as effective catalysts in ozonation processes of oxalic acid and dye C.I. Reactive Blue 5 [47]. Most of perovskites tested showed better performances in the catalytic ozonation of oxalic acid than single ozonation. On the contrary, in the case of the removal of dye, conversion values reached through single ozonation were slightly higher than in catalysed systems. However, the results in terms of TOC removal were better in the presence of catalyst and LaCoO3 allowed almost complete mineralization of dye after 3 h of reaction under the conditions tested. A key factor in the removal of contaminants seems to be the presence of lattice oxygen vacancies, which are able to activate adsorbed species.

In general, the catalytic activity seems to be enhanced by high surface area and easy access of reactants to the active sites. Then high surface area could be of interest to improve the activity of catalysts. Concerning to that, Afzal et al. [48] studied the behaviour of high surface area perovskites for catalytic ozonation of 2-chlorophenol. A nanocasting technique (NC), using SBA-15 as a template, was employed for the synthesis of NC-LaMnO3 and NC-LaFeO3 catalysts, with high surface area. Authors compared these catalysts with the same perovskites synthesized by conventional citric acid (CA) assisted route, CA-LaMnO3 and CA-LaFeO3, as well as with Mn3O4 and Fe2O3. They found that NC-perovskites, containing easily accessible active sites, showed higher catalytic activity (80% of TOC removal) than their counterpart CA-perovskites (35% of TOC removal). Mn3O4 and Fe2O3 were the worst catalysts.

Bromide, Br−, is usually present as micro-emerging pollutant in some water matrixes to be degraded. The ozonation treatment leads to the formation of bromate, BrO3 −, which is considered a potential carcinogen to humans by the World Health Organization, therefore being necessary its removal from the water. Zhang et al. [49] tested two perovskites, LaCoO3 and LaFeO3, as catalysts for the simultaneous removal of BrO3 − and benzotriazole (BZT) in the presence of ozone. 71% of BrO3 − was eliminated and BZT was completely degraded in only 15 min when LaCoO3 was used. LaFeO3 resulted in being inactive for BZT degradation; however, it reduced BrO3 − to HOBr/OBr− efficiently. The surface hydroxyl groups present in both perovskites were key in the involved reactions.

When the catalytic ozonation does not lead to a high degree of mineralization, it is necessary to combine it with other oxidation systems. The integration of catalytic ozonation and photocatalysis seems to be the most appropriate solution [50]. Photocatalytic ozonation is an advanced ozonation route allowing high removal of organic carbon by combining the beneficial effects of ozonation with the generation of hydroxyl radicals via electron hole formation (free radical oxidation). Considering that perovskites have been successfully used in catalytic ozonation, some authors studied the O3/UVradiation/perovskites system. Thus, Rivas et al. [51] carried out the advanced oxidation of pyruvic acid in presence of LaTi0.15Cu0.85O3 with several oxidation systems: O3, UV radiation, O3/UV radiation, O3/perovskite, UV radiation/perovskite, O3/UV radiation/perovskite, H2O2/UV radiation, H2O2/UV radiation/perovskite, being O3/UV radiation/perovskite the system investigated in more detail. The efficiency of the oxidation systems was examined in terms of the economic cost as a function of removal percentage of pyruvic acid and TOC. As expected, ozone was not able of eliminating pyruvic acid, displaying conversion values around 20% after 3 h of reaction. In contrast, application of UV radiation led to 40% of pyruvic acid elimination. Finally, for the combined O3/UV radiation/perovskite system, the pyruvic acid removal reached 100%, while the mineralization degree obtained was 80%. Therefore, under the operating conditions investigated, the photocatalytic ozonation seems to be the best option.
