*1.1.1 Mechanism 1: Decomposition of ozone by metal ions leading to the generation of free radicals*

The oxidation mechanism of organic compounds via ozonation is dependent on the pH of the reaction medium, (i) at basic pH ozone decomposes producing HO• radicals and other radical species in solution (Eqs. (7)–(9) and, (ii) at acidic pH, ozone is stable and reacts directly with organic substrates [31]. As it is well described in the literature, the generation of free radicals can subsequently oxidize the organic compounds more efficiently [23].

O3 þ OH� ! O2 •� <sup>þ</sup> HO2 • (7)

$$\text{O}\_3 + \text{O}\_2\text{"}^\text{-} \rightarrow \text{O}\_2 + \text{O}\_3\text{"}^\text{-} \tag{8}$$

$$\text{O}\_3\text{"}^\text{"} + \text{H}^+ \rightarrow \text{HO}\_3\text{"} \rightarrow \text{O}\_2 + \text{HO}\text{"}\tag{9}$$

The homogeneous catalytic ozonation occurs mostly at acidic pH values because at the real pH range for waters/wastewaters (6–8), the effect of metal ions is almost diminished [32].

In general, the mechanism of metal-catalyzed ozone decomposition with the generation of HO• radicals can be briefly expressed according to Eqs. (10)–(12) [31], being very similar to the Fenton process. The metal ions react with ozone or enhance its decomposition to generate HO• radicals and their regeneration occurs via the oxidation by HO2 �• radicals [9, 10].

$$\text{M}^{\text{n}+} + \text{O}\_{3} + \text{H}^{+} \rightarrow \text{M}^{(\text{n}+1)+} + \text{HO}^{\bullet} + \text{O}\_{2} \tag{10}$$

$$\text{O}\_3 + \text{HO}^\* \rightarrow \text{O}\_2 + \text{HO}\_2^{-\*} \tag{11}$$

$$\rm{M}^{(n+1)+} + \rm{HO}\_{2}^{-\*} + \rm{OH}^{-} \rightarrow \rm{M}^{n+} + \rm{H}\_{2}\rm{O} + \rm{O}\_{2} \tag{12}$$

The formation of HO• would be scavenged in the presence of excess metals (Eq. 13) [20, 21], so the optimization of catalyst dosage is also vital for catalytic ozonation process [23, 40].

$$\rm{M}^{n+} + \rm{HO}^{\bullet} \rightarrow \rm{M}^{(n+1)+} + \rm{HO}^{-} \tag{13}$$

One of the biggest challenges of this review was to find publications that represented the conditions found in the real waters (aquatic environments and wastewaters), for both the contaminants concentrations and pH. Since that, in the real waters, the contaminants concentrations range from ng/L and at the pH near neutral and most works present high values for them.

We gathered some works using metal ions for ozone catalysis and most of them report that Mn(II) and Fe(II) were the metals that showed the best results to increase ROS production. Sánchez-Polo and Rivera-Utrilla [37] tested the Mn(II) and Fe(II) ions as catalysts for the removal of 1,3,6-naphthalenetrisulfonic acid at acidic pH values; Xiao et al. [36] used the Mn(II) for the removal of 2,4-dichlorophenol; Ni et al. [32] used various metals for the removal of 2-chlorophenol at the acidic pH value and they found that Mn(II) was the most efficient catalyst tested. Okawa et al., [47] found that the presence of Fe(III) and Mn(II) enhanced the degradation of 2,4 dichlorophenol by ozone in acetic acid. However, Trapido et al. [48] observed no catalytic activity of Mn(II) for ozonation of dinitrobenzene. Wu et al. [31], Li et al. [49] and Ma & Graham [50] identified the optimal concentrations for metals ions to act as a catalyst for the decomposition of ozone into HO• radicals, in order to remove emerging contaminants, with an emphasis here on C.I. reactive red 2, alachlor and atrazine.

### *1.1.2 Mechanism 2: Complexes formation between organic molecule and the catalyst*

In this mechanism, the metal ions combine with organic molecules to form complexes, which are then oxidized by O3 and other oxidizing species [23].

Pines and Reckhow [35] reported that high mineralization of oxalic acid takes place via ozonation in the presence of Co(II) ions. This process was determined to have a high reaction rate, which increase with a decrease in pH. To prove that the oxidation reaction did not depend on the formation of HO• radicals, the researchers tested the reaction rate in the presence of tert-butanol, which is known as a HO• radical scavenger. Based on the results obtained, the authors confirmed that it is not changed in mineralization rates, proving that the HO• radicals was not responsible for mineralization of oxalate in the Co(II)/O3 system (**Figure 2**).

Soon after the work of Pines and Reckhow [35], Beltrán et al. [51] tested the same oxalic acid mineralization in the Co(II)/O3 system with the presence of tert-butanol and also confirmed that the HO• radicals were not responsible for mineralization of oxalic acid. Continuing their work with the mineralization of oxalic acid, Beltrán et al. [33] also found that Fe(III) ions act as a catalyst for the mineralization of oxalic acid in the same way as Co(II). In both cases (Co and Fe) ozone reacts with both negatively and positively charged complex moieties and HO• radicals are formed as secondary by-products. They proposed a sequence of reactions to explain the process (Eqs. (14)–(18)):

$$\text{Fe}^{3+} + \text{C}\_2\text{O}\_4 \text{ $^{-}$ } \rightarrow \text{FeC}\_2\text{O}\_4 \text{ $^{+}$ } \tag{14}$$

$$\text{FeC}\_2\text{O}\_4^{\cdot +} + \text{C}\_2\text{O}\_4^{\cdot -} \rightarrow \text{Fe(C}\_2\text{O}\_4)\_2^{-} \tag{15}$$

$$\text{Fe}(\text{C}\_2\text{O}\_4)\_2\text{ }^- + \text{C}\_2\text{O}\_4\text{ }^- \rightarrow \text{Fe}(\text{C}\_2\text{O}\_4)\_3\text{ }^{3-} \tag{16}$$

$$\text{FeC}\_2\text{O}\_4^+ + \text{O}\_3 \rightarrow 2\text{CO}\_2 + \text{Fe}^{3+} + 2\text{O}\_3^{\bullet -} \tag{17}$$

$$\text{Fe(C}\_2\text{O}\_4\text{)}\_2^- + 2\text{O}\_3 \rightarrow 2\text{CO}\_2 + \text{Fe(C}\_2\text{O}\_4)^+ + 2\text{O}\_3^{\bullet-} \tag{18}$$

$$\text{Fe}(\text{C}\_2\text{O}\_4)\_3^{3-} + 2\text{O}\_3 \rightarrow 2\text{CO}\_2 + \text{Fe}(\text{C}\_2\text{O}\_4)\_2^{-} + 2\text{O}\_3^{\bullet -} \tag{19}$$

Andreozzi et al. [52] explained the mechanism of the catalytic effect observed in Mn2+/O3 systems, as related to the formation of Mn(III) complex with oxalate ions (Ox). The molecular ozone attacked the oxalate ion radical and that leads to the formation of HO• radicals [53] (Eqs. (20)–(25)):

$$\text{Mn}^{2+} + \text{O}\_3 + 2\text{H}^+ \rightarrow \text{Mn}^{4+} + \text{O}\_2 + \text{H}\_2\text{O} \tag{20}$$

**Figure 2.** *Oxalic acid catalytic ozonation mechanism by means of the Co(II)/O3 system.*

$$\text{Mn}^{2+} + \text{Mn}^{4+} \rightarrow \text{Mn}^{3+} \tag{21}$$

$$\text{Mn}^{4+} + n\text{Ox}^{2-} \rightarrow \text{Mn}^{3+} \left(\text{Ox}^{2-}\right)\_{\text{n}} \tag{22}$$

$$\text{Mn}^{3+}\left(\text{Ox}^{2-}\right)\_{\text{n}} \rightarrow \text{Mn}^{2+}+\text{Ox}^{-}+(\text{n}-\text{1})\text{Ox}^{2-}\tag{23}$$

$$\rm Ox^{\star-} + O\_3 + H^+ \to 2CO\_2 + O\_2 + OH^\star \tag{24}$$

$$\text{OH}^\bullet + \text{Ox}^{2-} \to \dots \tag{25}$$

The transition metals are very important catalysts due to their characteristics [54]:


In summary, in mechanism 2, the ozone may equally efficiently attack neutral, positively and negatively charged metal-complex species, which could be a major reaction pathway for catalytic ozonation, especially for some low-molecular-weight acids, such as oxalic acid [35, 50].

### **2. Concluding remarks**

It is well described in the literature that catalytic ozonation in homogeneous phase is effective in removing a wide range of industrial effluents, products from the pharmaceutical industry, pesticides, and recalcitrant organics. However, homogeneous catalytic ozonation has the disadvantage of producing secondary contaminants from the addition of metallic ions [56]. In addition to the possibility of the residual concentration of metals exceeding regulatory limits for drinking water. Therefore, one more step in this system must be considered, to remove the metal ions from the treated matrices. This is the major drawback of applying homogeneous catalytic ozonation, especially when this process is applied for drinking water treatment. As a promising alternative to this inconvenience, heterogeneous catalysis appears which uses metals in the solid state (metallic oxides and metals on supports).
