From Fenton and ORR 2e⁻-Type Catalysts to Bifunctional Electrodes for Environmental Remediation Using the Electro-Fenton Process

Abstract

Currently, the presence of emerging contaminants in water sources has raised concerns worldwide due to low rates of mineralization, and in some cases, zero levels of degradation through conventional treatment methods. For these reasons, researchers in the field are focused on the use of advanced oxidation processes (AOPs) as a powerful tool for the degradation of persistent pollutants. These AOPs are based mainly on the in-situ production of hydroxyl radicals (OH^•) generated from an oxidizing agent (H₂O₂ or O₂) in the presence of a catalyst. Among the most studied AOPs, the Fenton reaction stands out due to its operational simplicity and good levels of degradation for a wide range of emerging contaminants. However, it has some limitations such as the storage and handling of H₂O₂. Therefore, the use of the electro-Fenton (EF) process has been proposed in which H₂O₂ is generated in situ by the action of the oxygen reduction reaction (ORR). However, it is important to mention that the ORR is given by two routes, by two or four electrons, which results in the products of H₂O₂ and H₂O, respectively. For this reason, current efforts seek to increase the selectivity of ORR catalysts toward the 2e⁻ route and thus improve the performance of the EF process. This work reviews catalysts for the Fenton reaction, ORR 2e⁻ catalysts, and presents a short review of some proposed catalysts with bifunctional activity for ORR 2e⁻ and Fenton processes. Finally, the most important factors for electro-Fenton dual catalysts to obtain high catalytic activity in both Fenton and ORR 2e⁻ processes are summarized.

1. Environmental Problems

One of the largest problems facing society today is water pollution, so special interest has been aroused in knowing the main pollutants that affect it and their subsequent treatment. Currently, some pollutants considered conventional (metals, fats and oils, nitrates, etc.) are controlled and monitored; however, more and more attention has been paid in recent years to those termed today as emerging [1,2,3]. Emerging contaminants can be defined as substances and/or chemical compounds that are not normally monitored but nevertheless remain continuous and remain present in water bodies despite being treated by conventional wastewater treatment plants [4,5,6,7].

Pharmaceuticals, along with their metabolites and degradation products, are considered emerging contaminants [8]. Their continuous discharge, together with their toxic and harmful levels, makes them highly dangerous for human and aquatic organisms [9,10]; therefore, it explains the importance of exercising controls and monitoring the concentrations of these pollutants both in the ecosystem and in effluents [11,12]. Currently, the global pharmaceutical market exceeds one trillion US dollars [13], which demonstrates the strong growth of this industry and its impact on the environment.

Among pharmaceutical products, antibiotics are of special interest because they are made to affect microorganisms, so they are prone to affect microbial communities in aquatic systems, causing a possible disappearance of them, which may have key functions in the ecosystem [14]. An antibiotic is defined as a chemotherapeutic agent capable of stopping the growth of microorganisms through cell destruction [15]. Antibiotic use has grown steadily globally, mostly in low- and middle-income countries [16]. It has been found that the presence of antibiotics in the environment favors the proliferation of microorganisms resistant to this type of drug [17].

Aquatic environments have been considered large deposits of antibiotics since this type of contaminant can arrive through two main ways, wastewater discharge and agricultural activity [18]. An example of this was a study conducted in 165 rivers in 72 countries, where researchers detected the presence of antibiotics in more than 60% of the 711 sampling sites [19]. In addition to these bodies of water, groundwater has also been susceptible to contamination with antibiotics mainly through the leaching of soils fertilized with manure from livestock, since it contains high concentrations of veterinary drugs [20].

Currently, wastewater treatment processes are composed of three phases in a general way:

• Primary treatment by physicochemical operations;
• Secondary treatment by biological processes;
• Tertiary treatment by additional processes.

However, these treatment technologies only achieve moderate to significant elimination of some antibiotics [21]. For this reason, other types of treatments have been explored to achieve almost total elimination of the antibiotics present in water. Currently, in this study, advanced oxidation processes (AOPs) are explored [22,23,24], such as Fenton, electro-Fenton, ozonation, photocatalysis, and electrocoagulation, among others [25,26,27,28,29], which are based on the in situ production of hydroxyl radicals (OH^•) generated from an oxidizing agent (H₂O₂, O₂) in the presence of a catalyst. OH^• has strong oxidizing power and thus, a high capacity to degrade organic compounds to simpler and less toxic biodegradable substances, and in some cases, even reaching complete mineralization [30].

2. Fenton Reaction

The first report on the Fenton reaction dates from 1876, where a mixture of H₂O₂ and Fe²⁺ salt (Fenton reagent) was used to achieve the oxidation of tartaric acid [31]; from there, various investigations have been carried out to apply this reaction in multiple applications. It is currently mainly used in environmental remediation [32,33,34,35,36].

The Fenton reaction can be considered an AOP, where it is based on a mixture of ferrous ion (Fe²⁺) (usually in a liquid state) with hydrogen peroxide (H₂O₂) [37], which, under the right reaction conditions, results in the formation of hydroxyl radicals [38] with high oxidative potential capable of degrading a wide range of contaminants (Figure 1) [39].

$${\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {{\mathrm{F}\mathrm{e}}^{3+}+\mathrm{O}\mathrm{H}}^{-}+{\mathrm{O}\mathrm{H}}^{•}$$

(1)

$${\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{F}\mathrm{e}}^{2+}{+\mathrm{H}}^{+}+{\mathrm{H}\mathrm{O}}_2^{•}$$

(2)

The Fenton reactions are based on the oxidation of Fe²⁺ to Fe³⁺, which simultaneously generates the hydroxyl radical (Equation (1)), followed by the reduction of Fe³⁺ to Fe²⁺ (Equation (2)), which allows the regeneration of the catalyst. However, the reaction in which OH^• is produced is much faster than that of catalyst regeneration [42], leading to the accumulation of Fe³⁺ and therefore decreasing catalytic efficiency.

Among its main advantages are its operational simplicity, zero toxicity, mild reaction conditions (ambient temperature and atmospheric pressures), and high degradation efficiency [43,44,45,46]. However, there are still some important limitations to overcome such as the slow reduction rate of Fe³⁺ to Fe²⁺ resulting in high levels of sludge generation with iron content, the deactivation of the catalyst, and a narrow pH working range [47,48,49,50,51].

The pH level is one of the most important parameters because it limits degradation efficiency. At pH values below 3, H₂O₂ goes on to form oxonium ions (H₃O₂⁺), which have less oxidative power than OH^•, while as the pH rises above 3, colloidal ferric species such as ferric hydroxide are formed from dissolved iron, which produces high amounts of sludge [52].

Recently, numerous investigations have been carried out (see Table 1) focused on the optimization of Fenton conditions. Variables such as catalyst pollutant and H₂O₂ concentration have been evaluated, as well as the pH of the solution, to determine under what conditions the highest percentage of pollutant molecule degradation is reached.

Typically, the pH is a crucial factor in the Fenton process, hardly affecting catalyst stability and pollutant degradation. From this, numerous investigations have been carried out with different pH values found at important degradation levels of pollutants such as antibiotics, colorants, and industrial wastewater. It is known that the pH optimal value for the Fenton reaction is closer to 3 due to the stability of H₂O₂, OH^•, and Fe³⁺. In this sense, Rao et al. [53] treated complex recalcitrant wastewater using FeSO₄·7H₂O as a catalyst (typical Fenton catalyst) in a concentration of 50 mg L⁻¹ and a pH value of 3 obtaining a degradation of 72% of DQO. On the other hand, Zhao et al. [54] removed a polymer quaternary ammonium salt using the same catalyst but using a pH of 2 and a Fenton catalyst concentration slightly lower (40 mg L⁻¹), achieving a percentage of elimination closer to 70%. Liu et al. [55] degraded a reference colorant for degradation tests, methylene blue, and analyzed the activation of the Fenton reaction under visible light irradiation using MoS₂-Fe (150 mg L⁻¹) as the Fenton catalyst at a pH value of 3, reaching a degradation of 65%. In turn, Qin et al. [56] achieved a 99.1% degradation level for a widely used antibiotic, amoxicillin, using 1 g L⁻¹ of a core-shell structured MnFe₂O₄@C-NH₂ at a pH of 3 (Figure 2).

Other antibiotics tested at pH 3 have been sulfamerazine (40 mg L⁻¹), sulfathiazole (0.5 mg L⁻¹), azithromycin (0.1 mg L⁻¹), norfloxacin (50 mg L⁻¹), and sulfamethoxazole (0.01 mg L⁻¹), with removal percentages >65%, using as Fenton catalysts, CNTs-Fe₃O₄ (0.5 mg L⁻¹), dissolved iron from zero-valent iron nanoparticles (9 mg L⁻¹), MnFe₃O₄-HS (1000 mg L⁻¹), Fe₃O₄/Schwertmannita/carbon (50 mg L⁻¹) (Figure 3) and fresh powder from Fe⁰, respectively [57,58,59,60,61]. On the other hand, Hommem et al. [62] studied a range of pHs from 3.5 to 4.5, for the degradation of amoxicillin (0.45 mg L⁻¹) using FeSO₄·7H₂O (0.095 mg L⁻¹) as catalysts, obtaining 100% degradation in all cases.

Other authors tried to increase the operating pH to facilitate the use of the Fenton process in a real application. Cheng et al. achieved a tetracycline hydrochloride degradation of 91.3% at a pH value of 4 using magnetic pullulan hydrogel as a Fenton catalyst (1250 mg L⁻¹) [63]. Similarly, Wang et al. [64] obtained a good tetracycline degradation (83%) at a pH value of 4.3 using iron loaded in graphitic carbon derived from microplastics. The highest pH values (5 to 5.5) tested also obtained good results: 100% degradation of carbamazepine and norfloxacin and 86.9% for ciprofloxacin using LaCu_0.5Mn_0.5O₃, FeSO₄·7H₂O with hydroxylamine (10 μM of Fe²⁺), and δ-Fe₃OOH/MWCNTs as Fenton catalysts, respectively [65,66,67]. However, neutralization of the final effluent is still required, so the search for catalysts and conditions that allow the development of this process at a neutral pH value is preferred. In this way, several researchers have studied the degradation of different pollutants such as amoxicillin, sulfadiazine, tetracycline, sulfadiazine, and rifampicin metronidazole, among others, working at pHs of 6 to 7.5, and obtaining degradation higher that 90% using goethite, Fe₂(SO₄)₃·H₂O, single-atom iron fixed in porous carbon, copper-modified MgFe-CO₃ double-layered hydroxide, rGO@nFe/Pd, magnetic biocarbon, CoFe₂O₄-S, and LaCu_0.8Mn_0.2O₃ [68,69,70,71,72,73,74,75] as Fenton catalysts. Guo et al. [76] also achieved the total degradation of ofloxacin at the widest pH using GO–Fe₃O₄ in a discharge plasma system. Wang et al. [77] also obtained the total degradation of ofloxacin using Fe₃O₄@S-doped ZnO at a pH of 5.2 to 9.0. For their part, Sun et al. [78] worked in a wide range of pH values from 3 to 9, degrading 99% of sodium sulfadiazine with FeCO₃ [6000 mg L⁻¹].

Table 1. Operating conditions for some Fenton processes.

Catalyst	Catalyst Concentration	H2O2 Concentration	Pollutant	Pollutant Concentration	Solution pH	Percentage of Elimination	Ref.
LaCu0.5Mn0.5O3	0.6 g L−1	0.7 g L−1	Carbamazepine (CZP)	15 mg L−1	5.5	100%	[65]
ZnO dopped Fe3O4@S	0.25 g L−1	5 mL L−1	Ofloxacin (OFX)	10 mg L−1	5.2–9.0	100%	[77]
FeSO4·7H2O	50 mg L−1	26.4 mM	Industrial wastewater (DQO)	850 mg L−1 (DQO)	3	72%	[53]
FeSO4·7H2O	40 mg L−1	23 mL L−1	Hydroxyethylpolydialyldimethylammonium-acrylamide-acrylic-acrylate (PDM)	200 mg L−1	2	69.3%	[54]
Siderite (FeCO3)	6 g L−1	100 mMol L−1	sodium sulfadiazine	50 mg L−1	9	99%	[78]
Goethite	-	460 mg L−1	Amoxicillin	105 mg L−1	6.5–7	83%	[68]
MnFe2O4@C-NH2	1 g L−1	3 mL L−1	Amoxicillin	30 mg L−1	3.0	99.1%	[56]
Fe2(SO4)3·xH2O	30 mg L−1	-	Tetracycline	50 mg L−1	6.0	90%	[69]
Fe2(SO4)·7H2O	0.095 mg L−1	2.35 mg L−1	Amoxicillin	0.45 mg L−1	3.5–4.5	100%	[62]
rGO–Fe3O4	0.23 g L−1	-	Ofloxacin	20 mg L−1	7.0–4.0	99.9%	[76]
MoS2-Fe	150 mg L−1	10 mMol L−1	Methylene blue	25 mg L−1	3	65%	[55]
CNTs–Fe3O4	0.5 g L−1	24.5 mM	Sulfamerazine	40 mg L−1	3	70%	[57]
Dissolved iron from zero valence iron nanoparticles (nZVI)	9 mg L−1	34 mg L−1	Sulfathiazole	0.5 mg L−1	3	>96%	[58]
Fe2(SO4)·7H2O with hydroxylamine	10 μM de Fe2+	1.0 mM	Norfloxacin	10 mg L−1	5	100%	[66]
CuS@Fe3O4/Pt	-	-	Tetracycline	40 mg L−1	-	78%	[79]
Single-atom iron fixed in porous carbon (Fe-ISA@CN)	100 mg L−1	10 mM	Sulfadiazine	2 mg L−1	6.5	96%	[70]
LDH–CuMgFe–CO3	0.5 g L−1	4 mMol L−1	Sulfathiazole	0.15 mg L−1	7.5	100%	[71]
rGO@nFe/Pd	200 mg L−1	167 mMol	Rifampicin	50 mg L−1	6.14	94.6%	[72]
MnFe2O4–HS	1 g L−1	29.4 mM	Azithromycin	0.1 mg L−1	3.0	92.6%	[59]
MagFePC (Hybrid)	250 mg L−1	5 mL L−1	Tetracycline	-	6.0	100%	[80]
Magnetic pullulan hydrogels	1.25 g L−1	1.25 mL L−1	Tetracycline hydrochloride	20 mg L−1	4.0	91.3%	[63]
Magnetic biocarbon	0.2 g L−1	50 mM	Metronidazole	20 mg L−1	6.0	97.4%	[73]
CoFe2O4–S	500 mg L−1	5 mMol L−1	Tetracycline	0.1 M	7.0	90%	[74]
COF (Photo-Fenton)	167 mg L−1	50 mL L−1	Rhodamine B	150 mg L−1	4.4	41.2%	[81]
Fe3O4/Schwettmann/Carbon	50 mg	45 μL	Norfloxacin	50 mg L−1	3.0	100%	[60]
δ-Fe3OOH/MWCNTs	235 mg L−1	20.6 mMol	Ciprofloxacin	10 mg L−1	5.3	86.9%	[67]
Fe2+	5.0 mMol L−1 de Fe2+	10 mmol L−1	Antibiotic resistance gene (ARG)	11.53 Log10(copies) g−1 dry sludge	3	66.86%	[82]
Fe–MPC	0.02 g L−1	1.0 mM	Tetracycline	40 mg L−1	4.3	83%	[64]
Fresh powder from Fe0	-	-	Sulfamethoxazole	10 μg L−1	3	100%	[61]
LaCu0.8Mn0.2O3	0.2 g L−1	13.8 × 10−3 mol L−1	Paracetamol	50 mg L−1	6.7	90%	[75]

Table 1. Operating conditions for some Fenton processes.

Catalyst	Catalyst Concentration	H2O2 Concentration	Pollutant	Pollutant Concentration	Solution pH	Percentage of Elimination	Ref.
LaCu0.5Mn0.5O3	0.6 g L−1	0.7 g L−1	Carbamazepine (CZP)	15 mg L−1	5.5	100%	[65]
ZnO dopped Fe3O4@S	0.25 g L−1	5 mL L−1	Ofloxacin (OFX)	10 mg L−1	5.2–9.0	100%	[77]
FeSO4·7H2O	50 mg L−1	26.4 mM	Industrial wastewater (DQO)	850 mg L−1 (DQO)	3	72%	[53]
FeSO4·7H2O	40 mg L−1	23 mL L−1	Hydroxyethylpolydialyldimethylammonium-acrylamide-acrylic-acrylate (PDM)	200 mg L−1	2	69.3%	[54]
Siderite (FeCO3)	6 g L−1	100 mMol L−1	sodium sulfadiazine	50 mg L−1	9	99%	[78]
Goethite	-	460 mg L−1	Amoxicillin	105 mg L−1	6.5–7	83%	[68]
MnFe2O4@C-NH2	1 g L−1	3 mL L−1	Amoxicillin	30 mg L−1	3.0	99.1%	[56]
Fe2(SO4)3·xH2O	30 mg L−1	-	Tetracycline	50 mg L−1	6.0	90%	[69]
Fe2(SO4)·7H2O	0.095 mg L−1	2.35 mg L−1	Amoxicillin	0.45 mg L−1	3.5–4.5	100%	[62]
rGO–Fe3O4	0.23 g L−1	-	Ofloxacin	20 mg L−1	7.0–4.0	99.9%	[76]
MoS2-Fe	150 mg L−1	10 mMol L−1	Methylene blue	25 mg L−1	3	65%	[55]
CNTs–Fe3O4	0.5 g L−1	24.5 mM	Sulfamerazine	40 mg L−1	3	70%	[57]
Dissolved iron from zero valence iron nanoparticles (nZVI)	9 mg L−1	34 mg L−1	Sulfathiazole	0.5 mg L−1	3	>96%	[58]
Fe2(SO4)·7H2O with hydroxylamine	10 μM de Fe2+	1.0 mM	Norfloxacin	10 mg L−1	5	100%	[66]
CuS@Fe3O4/Pt	-	-	Tetracycline	40 mg L−1	-	78%	[79]
Single-atom iron fixed in porous carbon (Fe-ISA@CN)	100 mg L−1	10 mM	Sulfadiazine	2 mg L−1	6.5	96%	[70]
LDH–CuMgFe–CO3	0.5 g L−1	4 mMol L−1	Sulfathiazole	0.15 mg L−1	7.5	100%	[71]
rGO@nFe/Pd	200 mg L−1	167 mMol	Rifampicin	50 mg L−1	6.14	94.6%	[72]
MnFe2O4–HS	1 g L−1	29.4 mM	Azithromycin	0.1 mg L−1	3.0	92.6%	[59]
MagFePC (Hybrid)	250 mg L−1	5 mL L−1	Tetracycline	-	6.0	100%	[80]
Magnetic pullulan hydrogels	1.25 g L−1	1.25 mL L−1	Tetracycline hydrochloride	20 mg L−1	4.0	91.3%	[63]
Magnetic biocarbon	0.2 g L−1	50 mM	Metronidazole	20 mg L−1	6.0	97.4%	[73]
CoFe2O4–S	500 mg L−1	5 mMol L−1	Tetracycline	0.1 M	7.0	90%	[74]
COF (Photo-Fenton)	167 mg L−1	50 mL L−1	Rhodamine B	150 mg L−1	4.4	41.2%	[81]
Fe3O4/Schwettmann/Carbon	50 mg	45 μL	Norfloxacin	50 mg L−1	3.0	100%	[60]
δ-Fe3OOH/MWCNTs	235 mg L−1	20.6 mMol	Ciprofloxacin	10 mg L−1	5.3	86.9%	[67]
Fe2+	5.0 mMol L−1 de Fe2+	10 mmol L−1	Antibiotic resistance gene (ARG)	11.53 Log10(copies) g−1 dry sludge	3	66.86%	[82]
Fe–MPC	0.02 g L−1	1.0 mM	Tetracycline	40 mg L−1	4.3	83%	[64]
Fresh powder from Fe0	-	-	Sulfamethoxazole	10 μg L−1	3	100%	[61]
LaCu0.8Mn0.2O3	0.2 g L−1	13.8 × 10−3 mol L−1	Paracetamol	50 mg L−1	6.7	90%	[75]

Apart from the commonly mentioned conditions, relevance has been found in some unusual factors, such as the slow and continuous addition of diluted H₂O₂, which improved the stochiometric efficacy of the reaction compared to the concentrated reagent being added only once [83]. Similarly, a high concentration of H₂O₂ causes an unfavorable effect on the reaction by bringing OH^• to H₂O [84], which decreases the efficiency of the process.

Although high percentages of elimination have been achieved for some pollutants, it is important to note that those tests in which dissolved iron has been used in solution (conventional process) have an optimal pH value close to 3, which maintains the previously mentioned limitations.

In order to overcome its limitations, one of the variations in the conventional process is the use of chelating agents, which can allow the reaction to happen at a pH close to neutral; however, the possible harmful effects such as toxicity and contribution to total organic carbon remain to limit its use [85]. Therefore, the use of heterogeneous catalysts is proposed, where the reactive species are generated mainly on the surface of the catalyst [86], and not, as in the homogeneous process, where the reaction happens in the solution. This makes the recovery and reuse of the heterogeneous catalyst easier [87] and decreases the leaching of Fe³⁺ and, therefore, the generation of sludge. However, some heterogeneous catalysts, although they present minor leaching, behave as if they were homogeneous [88].

An ideal Fenton heterogeneous catalyst must have high catalytic activity and stability, and be easy to recover and reuse [89,90]. To supply this, different materials have been evaluated with iron in different oxidation states. Fe⁰ has been reported as a possible heterogeneous Fenton catalyst when maintained in acidic conditions and in the presence of dissolved O₂ [91,92]. In the presence of oxygen, the in situ generation of H₂O₂ in zero-valent materials is possible, since the surface of the material can reduce O₂ using two electrons, accompanied by the generation of Fe²⁺ [93]. Iron oxides, possessing Fe in different oxidation states and being the most abundant minerals in the earth’s crust, have been evaluated in the Fenton process. Among these, goethite [94], hematite [95], and magnetite [96] have been investigated. Among the mentioned iron oxides, Fe₃O₄ has aroused greater interest, due to its zero toxicity, effective catalytic activity (thanks to its Fe²⁺ species in its crystal structure), as well as its easy recovery and reuse due to its magnetic properties [97,98,99,100]. However, Fe₃O₄ presents a low catalytic activity, mainly due to the slow speed of the Fe³⁺/Fe²⁺ cycle. Therefore, several strategies have been proposed to overcome this limitation, such as coating, use of materials as supports, adjuvants, etc. [101].

It has been found that the addition of different carbon materials can enhance the Fenton reaction because the presence of carboxyl, carbonyl, and quinone groups present in the carbon material surface accelerates the reduction of Fe³⁺ to Fe²⁺ (Figure 4) [102,103,104,105,106]. On the other hand, carbon materials have been also used as support of the Fenton catalyst. Sun et al. [107] prepared ferric hydroxyquinoline supported on active carbon fiber, which enhanced the Fe³⁺/Fe²⁺ cycle through the free electrons present in the carbon material. Yao et al. [108] supported copper ferrite on reduced graphene oxide finding that the presence of two metal ions improved the redox cycle of both ions Cu and Fe together, with a synergistic effect between the nanocube structure of the copper ferrite and the mesoporosity of the support. Kuśmierek et al. [109] and Cruz et al. [110] used organic xerogel as support for Fe/N and CoFe₂O₄ and, in both investigations, a catalytic improvement was observed and ascribed to a higher activation of H₂O₂ due to the coexistence of two differents atoms (Fe and N or Co and Fe) in a same matrix.

This bimetallic synergism and the effect of the support were also observed by other authors. Zhang et al. [111] prepared Fe/Cu bimetallic-phase-supported montmorillonite using a two-step strategy of impregnation and calcination and found excellent efficiency in the removal of rhodamine B in comparison to single-metal-loaded montmorillonite. These results were ascribed to (i) the expansion of the montmorillonite pore size due to the insertion of Cu and Fe atoms making this clay more suitable as a Fenton catalyst, (ii) the electrons involved in the redox reaction of bimetallic active sites promote the rate of mutual oxidation reactions by enhancing the H₂O₂ conversion into oxidant radicals, and (iii) the leaching of the pure metal ions decreases due to the good stability of support. Li et al. [112] synthesized Fe–Mn/SAPO-18 zeolite catalysts by ion exchange, maintaining the zeolite structure. The author observed an enhanced degradation efficiency of methyl orange compared to the single-metal catalysts (Fe and Mn), which was also attributed to the synergistic effect of bimetal that improves OH^• radical generation and also promotes the formation of another active species, HO₂^•. Similarly, Xu et al. [113] used sepiolite as support for Fe₃O₄ dispersion to be used as catalysts for the Fenton degradation of bisphenol A (BPA). The high surface area of the catalyst and the electron transfer and π-π interactions are responsible for the good BPA adsorption properties. Thus, the preconcentration of BFA by adsorption near the active site centers and the generation of hydroxyl radicals in the reaction of H₂O₂ and active sites make the Fenton degradation easier on the surface of Fe₃O₄-Sep. Ferroudj et al. [114] tried to analyze the catalyst’s structure–activity relationships and the effect of Fe³⁺ doping in the Fenton degradation of methyl orange (MO) and reactive back 5 (RB5) by the synthesis of Fe³⁺ doped and undoped maghemite/silica microporous and mesoporous microspheres (γ-Fe₂O₃/SiO₂ MS). For that study, the authors concluded that porosity and Fe dispersion play an important role in the Fenton reaction. The creation of mesoporosity in the support enhances the reactants and products diffusion allowing a faster degradation of the tested dyes regarding its microporous counterpart, mainly for the RG5 molecule due to the biggest size. This degradation activity was improved after the doping of the silica framework with Fe³⁺. Thus, the best results were obtained for the Fe-doped γ-Fe₂O₃/SiO₂ MS mesoporous catalyst. This material allows an almost complete degradation of MO in 25 min in contrast to the doped microporous catalyst, which achieved 92% of MO degradation in 1 h, and the undoped materials, which produces a much low degradation efficiency (63%) in 1 h. Similar results were obtained in the degradation of RB5.

The above-commented strategies to enhance the Fenton active-phase performances and the improvements achieved are collected and summarized in

Table 2. Strategies and improvements in the development of electro-Fenton Catalysts.

Catalyst	Strategy	Improvement	Ref.
Fe (NO3)3·9H2O	Addition of MWCNT and complexation on the surface of the MWCNT through the surface groups.	The reaction rate was 23 times higher than in traditional Fenton. Faster Fe3+/Fe2+ cycle (Demonstrated by the generation of OH• and HO2•, combined with an accelerated decomposition of the contaminant). Better catalytic activity by MWCNT carboxylates groups. Accelerating the intermediate step from Fe-OOH to Fe2+.	[102]
FeCl3·6H2O	Addition of carbon materials (powdered activated carbon and carbon nanotubes).	The reduction from Fe3+ to Fe2+ was accelerated. Due to the reducing power of carbon materials (especially carbonyl/quinone groups). Carbon nanotubes had a lower rate of decomposition of H2O2 than powdered activated carbon.	[103]
Fe (NO3)3·9H2O	Addition of hydrothermally prepared carbon.	Promoted the Fe3+/Fe2+ cycle through the transfer of electrons from carbonaceous material to Fe3+. Hydroxyl groups on the surface of hydrothermal carbon strongly affect the reduction of Fe3+.	[104]
Fe2O3@FeB	FeB coating to Fe2O3 core.	An efficient Fe3+/Fe2+ cycle was achieved, thanks to the oxidizing and constantly ceding electrons to Fe3+. The Fe2+ could also generate H2O2 in situ by the activation pathway of a single electron of O2.	[115]
FeCl3·6H2O	Addition of granular activated carbon.	The formation of Fe2+ was accelerated, which improved the catalytic oxidation of the target contaminant.	[105]
Ferrihydrite	Addition of biocarbon and S. oneidensis.	Faster Fe3+ reduction rate.	[106]
Montmorillonite (Fe/Cu-MMT)	Preparation of Fe–Cu bimetallic active sites intercalated on montmorillonite.	The bleaching efficiency of the contaminant was higher, thanks to the electrons involved in the redox reaction of both metals, which promoted the rate of mutual oxidation. A decrease in the leaching of pure metal ions of Fe and Cu, by presenting good stability with the support.	[111]
Fe–Mn/SAPO-18	Fe–Mn bimetallic active sites supported on SAPO-18.	Improved degradation efficiency of the contaminant, compared to the single-metal catalysts (Fe and Mn) supported on the zeolite.	[112]
Fe3O4–Sep	Sepiolite support for Fe3O4 nanoparticles.	Improved adsorption of the catalyst and, therefore, the in situ degradation of the contaminant on the surface of Fe3O4-Sep.	[113]
γ-Fe2O3–SiO2	Mesoporosity creation and Fe3+ doping of maghemite/silica microspheres.	The substitution of microporosity or mesoporosity led to a greater discoloration of the effluent to be treated.	[114]
QuFe@ACF	Active carbon fiber support for ferric hydroxyquinoline.	Free electrons in the carbon fibers were able to drive electron transfer to ferric 8-hydroxyquinoline to promote the Fe3+/Fe2+ cycle faster.	[107]
Fe-C	Fe-doped carbon xerogel.	Improved dispersion of Fe0 species.	[116]
CF/rGO	Reduced graphene oxide support for copper ferrite.	The redox efficiency improved due to the two metal ions (Cu2+/Cu+ and Fe3+/Fe2+). Synergistic effect between the nanotube structure of the copper ferrite together with the mesoporosity of the support.	[108]
Fe–N/C	Carbon xerogel support for heteroatoms (Fe, N).	Improved mass transfer to catalytic active sites. Increased activation of H2O2 due to the coexistence of Fe and N.	[109]
CoFe2O4/NOM	Xerogel support from the natural organic matrix (NOM) for CoFe2O4.	Increased surface area, increasing exposure to catalytically active sites. Improved electron transfer between metal ion pairs, which increased the speed of the redox cycle (Co2+/Co3+ and Fe2+/Fe3+).	[110]

.

As evidenced in the literature, supporting Fenton heterogeneous catalysts is an effective method to increase the speed of the Fe³⁺/Fe²⁺ cycle, making it possible to enhance the efficiency of the process. In the same way, an interesting option is the use of metal-doped (e.g., N–Cu) or bimetallic active phases (e.g., Fe–Mn or Fe–Cu), which not only favor redox or Fenton reactions but also more complex reactions such as the oxygen reduction reaction (ORR) (e.g., nitrogen), which is very important for the electro-Fenton process, another variable of conventional Fenton that is addressed below.

3. Electro-Fenton Process

Heterogeneous Fenton has been proven to be a process with considerable potential for treating a wide range of pollutants. However, the storage, transport, and handling of H₂O₂, as well as the high costs associated with the excess consumption of the reagent are a limitation of the process [117]. To overcome these problems, the heterogeneous electro-Fenton (HEF) has been proposed, where the main contributions are the in situ production of H₂O₂ through the ORR at the cathode, the reduction of Fe³⁺ in the same electrode to supply Fe²⁺ again and the improved generation of OH^• by the applied electricity [118,119,120].

This process has been successfully used for the degradation of different recalcitrant pollutants such as drugs, dyes, petroleum derivatives, etc. [121,122,123,124]. The mechanism of the electro-Fenton process is given by several stages, summarized in [125]:

• Electrogeneration of H₂O₂ (cathodic reduction of O₂).
• Production of OH^• (Fenton reaction).
• Promotion of the formation of OH^• physisorbed on the surface of the electrode.
• Regeneration of Fe²⁺ through the direct reduction of Fe³⁺ at the cathode.

In general, the reactions involved in the process can be summarized in Equations (3)–(6) [126,127,128]. However, it is necessary to pay special attention to the possible side reactions associated with the electrolysis of water that can decrease the performance of the catalytic activity of the process [129].

$${\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{e}}^{-}\to {\mathrm{F}\mathrm{e}}^{2+}\left(\mathrm{C}\mathrm{a}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{d}\mathrm{e}\right)$$

(3)

$${\mathrm{O}}_2+{2\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2{\mathrm{O}}_2\left(\mathrm{C}\mathrm{a}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{d}\mathrm{e}\right)$$

(4)

$$2{\mathrm{H}}_2\mathrm{O}\to {4\mathrm{H}}^{+}+{\mathrm{O}}_2+4{\mathrm{e}}^{-}\left(\mathrm{A}\mathrm{n}\mathrm{o}\mathrm{d}\mathrm{e}\right)$$

(5)

$${\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{H}}^{+}\to {\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}\mathrm{H}}^{•}$$

(6)

One of the most important benefits of this process to highlight is the expansion of the working pH range with the removal of contaminants at a pH close to neutral as well as its improvement in levels of biodegradability and toxicity [130,131]. However, as evidenced by Equations (4) and (6), ORR is the limiting step of the process, so it is important to pay attention to the cathode used [132,133,134].

Traditionally, carbon materials have been used as cathodes due to their good conductivity, stability, and low cost. Thus, these carbon materials are presented as excellent candidates as high-performance cathode electrodes for electro-Fenton reactions for several reasons that are described in the following. Sun et al. [135] synthesized a family of porous biochars by the chemical activation of cellulose powder with ZnCl₂ and demonstrated that the carbonization temperature plays an important role in the catalytic performance in the EF oxidation of organic contaminants. While a high carbonization temperature favors the generation of water via the 4-electron pathway, a mild temperature of carbonization (400 °C) improved the generation of H₂O₂ via the 2-electron pathway. This different behavior is attributed to the surface chemistry of the carbon surface; mild carbonization temperatures provide an oxygen-functional-group-rich surface with a high C-O/C=O ratio, which enhances the H₂O₂ selectivity. In contrast, the use of higher carbonization temperatures (700 °C) guarantees a highly graphitic structure and created a carbon defect-rich porous structure highly selective to the 4-electron route. As a consequence, the biochar prepared at 550 °C presents a better catalytic performance as the cathode material for the EF degradation of several organic pollutants in presence of Fe (II), presenting a high H₂O₂ production rate (796.1 mg/g/h at −0.25 V vs. RHE at pH of 1) and achieving an almost complete removal (99%) of phenol after 1 h (at −0.25 V vs. RHE, pH = 3, [Fe (II)] = 0.5 mM, electrolyte: ((Na₂SO₄) = 0.1 M). Ergan et al. [136] also analyzed the effect of the thermal activation of activated carbon fibers in the H₂O₂ electrogeneration and EF efficiency. In this case, the effect of the carbonization time (from 1 h to 10 h) at 900 °C and also the carbonization atmosphere (CO₂ or N₂) was analyzed, observing that the best H₂O₂ production and dye (acid orange 7) degradation (92.9%) was obtained when the fibers were activated in an N₂ atmosphere for 5 h due to a superior hydrophobicity, crystallite size, and larger electroactive surface area. Ramírez-Pereda et al. [137] compared two types of carbon materials: carbon fibers (CF) and vitreous carbon (VC) as electrodes for the electrogeneration of H₂O₂. The authors found that the generation of H₂O₂ is enhanced using CF (96% of H₂O₂ production efficiency) in comparison to VC (54% of H₂O₂ production efficiency), which was attributed to possible competitive reactions. Xia et al. [138,139] analyzed the performance of a polyacrylonitrile-based carbon fiber brush (PAN–CFB) as a cathode electrode for ORR reduction to H₂O₂ and studied the effect of the electrochemical modification of its surface. For the unmodified PAN–CFB, the highest production of H₂O₂ (current efficiency > 90%) was observed at a potential of −0.9 V in an acidic medium, whereas for electrochemically modified PAN–CFB, it was obtained at −0.4 V. However, in a neutral medium (pH = 7), an electrochemically modified PAN–CFB cathode inefficiently generated H₂O₂ (41.2% of current efficiency) in comparison with unmodified PAN–CBF, which still retains the high current efficiency (>90%). This was explained based on the surface chemistry modification after the electrochemical modification treatment, which converts the N-containing groups from pyridinic-N to 2-pyridone—these newly ORR active sites were influenced by solution pH, leading to different pathways in acidic or neutral media. Gao et al. [140] synthesized porous carbon felt modified with polypyrrole (Ppy) and anthraquinone 2-sulfonate, which had a higher specific surface area with more active sites for ORR; in addition, the modification with Ppy decreases the charge transfer resistance. In turn, Chen et al. [141] showed that the adsorption capacity of the carbon-based cathode plays an important role in the EF degradation of tetracycline. For that, they activated carbon felt with KOH at different temperatures, finding that the surface area and the oxygen-containing functional groups on the surface increased in the KOH activation, which significantly enhanced the drug adsorption capacity (being the sample carbonized at 900 °C, which presents the highest adsorption capacity) and also produced the best tetracycline degradation. This highest degree of pollutant degradation (tetracycline 80% removal) was explained based on the 3D structure and good adsorption capacity of activated carbon felt, which creates a space where the organic pollutant is adsorbed and preconcentrated near the active sites in charge of generating the H₂O₂, and, consequently, exposed to a high concentration of hydroxyl radicals in the system of homogeneous catalysis. Wang et al. [142] obtained a complete degradation of dimethyl phthalate using KOH-activated graphite felt at 900 °C. This high performance was attributed to the high surface area, the presence of oxygenated functional groups, and the more hydrophilic surface generated by the activation of the KOH.

Want et al. [143] pointed out the importance of electrode conductivity on its EF performance. For that, CNT and graphene-modified carbon felt were synthesized and the effect of this modification was analyzed in the degradation of azo-dye (RB5) by EF reaction. The modification of the carbon felt with CNT and graphene increases the degradation rate by 1.2 and 1.5 times than with the original carbon felt, respectively, giving degradation rates of 55.3% and 70.1%, respectively, which was attributed to the improved roughening of the smooth carbon felt fiber (specific surface area), conductivity, and corrosion resistance.

Since the ORR reaction is crucial for the H₂O₂ production, the improvement in the O₂ supply to the active sites seems to be a key factor for the improvement in the EF efficiency, together with, as described above, the preconcentration of the pollutant near the H₂O₂-producing active sites. In this sense, Özcan et al. [144] compared carbon felt with carbon sponge as cathode materials for the degradation of basic blue 3 (BB3) by EF process finding a faster BB3 degradation using the carbon sponge-based electrode, which was associated with the major H₂O₂ production. Sopaj et al. [145] also evaluated the performance of carbon sponges in comparison with carbon felt and stainless steel in sulfamethazine degradation. The highest production of H₂O₂, and, consequently, the highest degradation, was obtained using the carbon sponge instead of carbon felt, the classical cathode for the electro-Fenton process, and stainless steel. This improved catalytic performance of carbon sponge is attributed to the high specific surface are and porosity, which results in inadequate conditions for the mass transport of species of interest (O₂ and H₂O₂) throughout its 3D structure. Ganiyu et al. [146] also synthesized a carbon foam with high performance (total degradation and high mineralization) in the EF degradation of sulfanilamide, which was also attributed to the interconnected spherical cells’ porous structure (Figure 5), with high-area B.E.T. that improved the diffusion of substances and gaseous oxygen in the pores and active sites of the cathode, together with an excellent hydrophobicity and conductivity. Nonetheless, it exhibited poor performance in the electro-regeneration of Fe²⁺ from the reduction of Fe³⁺. This poor Fe²⁺ electro-regeneration of “gas diffuser materials” was also observed by Brillas et al. [147]. They demonstrated that H₂O₂ generation is important; however, it is not a uniquely important factor in the degradation of pollutants by EF. They analyzed the Fe(III)–EDDS-assisted EF degradation of butylated hydroxyanisole (BHA) using a carbon felt or an air-diffusion electrode as a cathode. They observed a much higher Fe(III) reduction efficiency in the novel Fe(III)–EDDS-assisted EF process in comparison with conventional EF. Moreover, the carbon felt resulted in a higher butylate hydroxyanisole degradation despite having an H₂O₂ generation lower than the air-diffusion cathode, which was attributed to the greater regeneration of Fe²⁺ observed using the carbon felt. To overcome this limitation, Chu et al. [148] used a dual cathode system: a gas diffusion electrode (GDE) to generate H₂O₂ by O₂ reduction and a graphite electrode for the regeneration of Fe²⁺ from Fe³⁺ and analyzed its performance as a conventional single-cathode system in the degradation of 4-nitrophenol. They observed that the dual cathode system led to a more effective degradation of 4-nitrophenol, even with a lower initial Fe²⁺ concentration attributed to the rapid change between the Fe³⁺/Fe²⁺ couple in the dual cathode system. Since the main function of GDE is the electrogeneration of H₂O₂, the regeneration of Fe²⁺ by the cathodic reduction in the single-cathode system is very weak, whereas this regeneration was enhanced by the reduction at the graphite cathode in the dual-cathode system. Chu et al. [149] have also improved the reduction of Fe³⁺ of graphite by its modification with acidified carbon nanotubes, obtaining a degradation of 94.7% of p-nitrophenol. The acidification of the carbon nanotube enhanced the performance in the Fe³⁺ reduction in comparison with the unmodified graphite cathode.

Other advanced electrodes that improve the O₂-supply are gas diffusion electrodes (GDE) evaluated by different researchers. Zarei et al. [150] synthesized a three-dimensional graphene (3DG)-based carbon paper GDE for the removal of nalidixic acid. The nalidixic acid removal was 90%, which was attributed to a higher surface area with more irregularities and roughness than graphene oxide, allowing rapid diffusion of O₂ to achieve higher H₂O₂ production. On the other hand, Zhou et al. [151] used graphite felt, but in this case, as a floating cathode in which one side of a porous cathode is open to the air, and the other side is submerged in the aqueous solution instead of the traditionally submerged cathode. This novel strategy allows the creation of a three-phase boundary (O₂ gas, electrolyte, and cathode) allowing the usage of O₂ from both air and the electrolyte. This disposition is more energy-efficient and cost-effective in comparison with a GDE because avoids hydrophobic gas diffusion layer formation and the use of an air or O₂ pump. An ibuprofen degradation of 78.3%, with only 25.4% obtained in the conventional position. Liu et al. [152] used a GDE of graphene doped with N-doped carbon nanotubes as the cathode in the dimethyl phthalate removal (100%), finding that the graphene–nanotubes interaction significantly improve the ORR activity.

The hydrophobicity of the cathode ©s also so important in the EF reaction. Chu et al. [153] managed to completely degrade cefepime with a sandwich-like superhydrophobic carbon cathode composed of graphite, carbon nanotubes, and PTFE observing that the hydrophobicity of the electrode significantly affects the ORR to H₂O₂ (Figure 6). Moreover, the prepared superhydrophobic carbon cathode resulted in an excellent reduction of O₂ to H₂O₂, and, consequently, an excellent cefepime degradation, which is mainly dependent on the ∙OH production via Fenton reactions. Karatas et al. [154] also attributed the excellent performance of a carbon black electrode for the full degradation of atrazine to the superhydrophobic character of the electrode obtained because this high hydrophobicity provides high H₂O₂ yield by supporting the mass transfer of oxygen molecules.

Another strategy to improve EF performance is the use of metallic-doped electrode materials. Huang et al. [155] prepared a graphite felt-based cathode by first doping the graphite felt with Fe–Mn, then with active carbon, carbon black, and PTFE, and used it for the degradation of ciprofloxacin (Figure 7). This modified graphite felt-based cathode showed a higher degradation efficiency (95.4%) and wider operational pH range than their unmodified counterpart, which was principally associated with the larger surface area and volume pore as well as more active sites. Various researchers have also proposed doping carbon materials for improving EF efficiency. Ma et al. [156] doped multiwalled carbon nanotubes with Ag and Cu foam and noted that the high Ag dispersion improves the ORR and Fe²⁺ regeneration. Dung et al. [157] synthesized cobalt ferrite-coated carbon felt for the degradation of tartrazine (97%), where it is noted that the presence of Co³⁺/Co²⁺ and Fe³⁺/Fe²⁺ pairs improve the pollutant degradation; additionally, an additional Fenton catalyst was not needed in this test because the cathode fulfilled both functions. Paz et al. [158] used doped Vulcan XC72 with tungsten oxide nanoparticles and showed an improvement in the H₂O₂ production due to the surface being more acidic and hydrophilic. Fdez-Sanromán et al. [159] degraded 100% of pymetrozine with a cathode based on carbon felt and modified with carbon nanofibers doped with iron; the improvement was explained by the porous structure that was enhanced with the incorporated nanofibers, and it was also explained that the catalytic activity of a typical Fenton catalyst is more effective in the presence of iron.

In general, it could be summarized that the important parameters to take into account in the cathode preparation are the mesoporous structure that translates into the high surface area, the hydrophobicity that plays an important role in terms of the diffusivity of O₂ inside and outside of the electrode, as well as the regeneration of Fe²⁺ that happens in the electrode, and finally, the composition of the cathode, especially the presence of efficient nitrogenous groups toward ORR [160,161]. However, more in-depth research should be conducted, despite the extensive literature that supports the above. In this regard, Cordeiro stressed that graphite within a series of carbonaceous materials presented the smallest B.E.T. area and zero existence of oxygenated groups on the surface, and still showed the best efficiency of H₂O₂, which was associated with its laminar structure and low hydrophobicity leading to the faster release of H₂O₂ compared to more porous materials [162].

Another important parameter to consider in the cathode environment is the balance between the three reactants (oxygen, proton, and electron), which must have a molar rate of 2:1:2, respectively [163]. For this, it is important to start studying how to supply air/O₂, which is based on two types of aeration specifically.

• Direct aeration in solution.
• Aeration through a gas diffusing electrode (GDE).

Direct aeration in solution through bubbling is the traditional method; on the other hand, GDEs have the advantage that thanks to their porous structure and coexistence of the limits of three phases, they have solved the problems of low solubility and mass transfer of O₂ [164], which can accelerate ORR kinetics and gas-use efficiency, summarizing even better energy use. In general, GDEs can improve ORR mainly due to two main mechanisms [165].

• The increase in the mass transfer of O₂ improves the contact of the gas with the catalyst and, therefore, the active sites for the ORR.
• The increase in the mass transfer is not only in reference to O₂ but also between the catalyst and electrolyte, which promotes the release of H₂O₂ and active sites, improving the formation of the oxidizing species.

GDEs can be defined as porous electrodes in which the solid part is in contact with the gas phase, and then with the solution to be treated [166], which generates a three-phase interface in which the reactions of interest will happen. GDEs are mainly composed of a catalyst layer (in contact with the solution) and a diffuser (in contact with air or oxygen) (Figure 8) [167]. At this point, it is important to clarify that a very thick layer of catalyst can decrease the output rate of H₂O₂ to the solution, which can cause it to degrade on the surface of the material. An alternative to adjusting the thickness of the layer or avoiding its limitation is to adjust the hydrophobicity of the catalyst [168].

As it has been pointed out, good results have been obtained in the EF degradation of a huge variety of emerging pollutants. However, since H₂O₂ production by the O₂ reduction at the cathode seems to be the limiting step, a deeper investigation into ORR is necessary to better understand how cathode variations benefit or affect the efficiency of the EF process.

4. Promising New Fenton Catalysts: Spinel Ferrites and Perovskites

Fenton catalysts with higher reaction rates are currently being sought to increase their overall efficiencies. Among these, spinel ferrites have attracted increasing interest in recent years due to their low cost, excellent catalytic activity, and magnetic properties that allow a facile separation and reutilization. In this section, the synthesis routes, advantages, and applications of spinel ferrites will be described.

There are different methods of synthesis within which the following stand out:

• Chemical coprecipitation

Chemical coprecipitation is one of the most promising methods thanks to its ease of realization, as well as the possibility of mass production. It is mainly based on the co- precipitation of materials by a pH change. In general, this method consists of a mixture of precursors with the species of M²⁺ and M³⁺ (metal), to later precipitate the particles by the addition of a base (KOH, NaOH, or NH₄OH). It can be summarized in Equation (7) [169,170,171,172].

$${\mathrm{M}}^{2+}+2{\mathrm{F}\mathrm{e}}^{3+}+{8\mathrm{O}\mathrm{H}}^{-}\to {\mathrm{M}\mathrm{F}\mathrm{e}}_2{\mathrm{O}}_4+4{\mathrm{H}}_2\mathrm{O}$$

(7)

4.1. Sol–Gel Method

This method is one of the most used for the synthesis of nanoparticles due to its relatively low cost, operational simplicity, and high homogeneity of the material obtained. It is based on a series of reactions, summarized in the hydrolysis of a metal alkoxide and the subsequent polycondensation of the hydroxyl groups formed, which produces a three-dimensional matrix [173,174,175], which must subsequently be subjected to thermal processes to improve the crystallinity of the nanoparticles.

4.2. Hydrothermal/Solvothermal

This process can be defined as a series of chemical reactions that occur in a closed system with one or more precursors in the presence of the solvent (water for the hydrothermal case), at a temperature above the boiling point of this [176]. This process has several advantages such as its operational simplicity, versatility, and low cost, among others [177,178]. An important advantage to highlight is the well-controlled diffusivity within the system [179], which allows good control of the structure and morphology of the synthesized particles.

From these synthesis methods, it has been possible to obtain different new Fenton-type catalysts at the nanometer scale, with different coatings and supports [180,181,182] in order to improve the efficiency of the hetero-Fenton process. It has been established that in the hetero-Fenton process, having a catalyst with a higher amount of transition-metal ions and a higher specific area leads to better Fenton activity [183,184]. Fe₃O₄ is currently being extensively investigated as a Fenton catalyst [185,186,187] due to its relatively high activity and easy magnetic separation. An essential advantage is the presence of Fe²⁺ and Fe³⁺ in a single material, due to its cubic structure where half of the Fe³⁺ ions occupied all of the tetrahedral sites and the Fe²⁺ ions are founded in the octahedral, allowing the presence of the two important species in Fenton processes. It has been reported in the literature that transition-metal ions occupying tetrahedral sites are catalytically inert, while those located at octahedral sites tend to determine catalytic activity, which is due to effect that the metal cations in these positions are found exclusively on the surface of the solid and thus take part in reactions to generate OH^• radicals [188]. It has been suggested that the activation of H₂O₂ in the presence of magnetite takes place on the surface of the solid, i.e., the Fe³⁺-OH [189].

According to crystalline field theory, common metal cations have the following order of preference to occupy octahedral sites, Cr³⁺ > Ni³⁺ > Cu³⁺ > Al³⁺ > Mg²⁺ > Fe²⁺ > Co²⁺ > Fe³⁺ > Mn²⁺ > Zn²⁺ [190]. The most studied are Fe, Mn, Cu, Ni, Zn, and Co [191,192].

Spinel ferrites have a general formula of A²⁺B³⁺₂O⁻²₄ and a cubic lattice structure, where positions A and B are occupied by divalent and trivalent metal cations, according to the distribution of cations in the lattice, and the spinels can be normal, random, or inverse. In a normal spinel, tetrahedral sites are related to position A and octahedral sites to B. In a reverse spinel, tetrahedral sites are occupied by half of the B cations and the octahedral sites by all of the A cations. This distribution of cations is responsible for determining the magnetic properties of ferrites [193]. There are different types of spinel-structured materials; however, in this work, we will focus specifically on three of them: copper and cobalt ferrites and iron cobaltite.

Copper ferrite (CuFe₂O₄) is a well-known material with a reverse-spinel structure, which has a stable structure that reduces metals leaching, together with unique magnetic, electrical, physical, and chemical properties [194,195,196,197]. These properties make them a promising Fenton-type catalyst. Cu⁺ ions have been reported to have the ability to generate OH^• radicals by a mechanism similar to that of Fe²⁺ according to Equations (8)–(10) [198].

$${\mathrm{C}\mathrm{u}}^{+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{C}\mathrm{u}}^{2+}+{\mathrm{O}\mathrm{H}}^{-}+{\mathrm{O}\mathrm{H}}^{•}$$

(8)

$${\mathrm{C}\mathrm{u}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{C}\mathrm{u}}^{+}+{\mathrm{H}}^{+}+{{\mathrm{H}\mathrm{O}}_2}^{•}$$

(9)

$${\mathrm{C}\mathrm{u}}^{2+}+{{\mathrm{H}\mathrm{O}}_2}^{•}\to {\mathrm{C}\mathrm{u}}^{+}+{\mathrm{H}}^{+}+{\mathrm{O}}_2$$

(10)

Feng et al. [199] synthesized CuFe₂O₄ nanoparticles to be used as a Fenton catalyst in the degradation of sulfanilamide. They suggest that CuO is more reactive and effective than Fe³⁺ for the activation of H₂O₂ and, which is more important, can work in a higher pH range than conventional iron oxides. Moreover, Zhang et al. [200] observed that the leaching of Cu²⁺ is 30 times lower in CuFe₂O₄ than with CuO. The promising performance of the copper spinel ferrite as a Fenton catalyst was pointed out by several authors. Suraj et al. [201] synthesized CuFe₂O₄ by the chemical coprecipitation method and used it as a heterogeneous Fenton catalyst for the treatment of pulp and paper wastewater, obtaining a 78% elimination of the chemical oxygen demand. Ding et al. [202] demonstrated that the morphology of the spinel is also very important. They synthesized hollow CuFe₂O₄ spheres with oxygen vacancies (Figure 9), which demonstrated greater degradation of ciprofloxacin than normal CuFe₂O₄ particles. This better performance of hollow spheres was attributed, among other factors, to the synergistic oxygen vacancies and confinement effects on the catalyst surface. The oxygen vacancies produce highly active electron-rich Cu⁺ species, which enhanced the H₂O₂ activation and, thus, the hydroxyl radical generation. In turn, the hollow structure is responsible for concentrating the organic pollutants near the ^•OH-generator active sites, improving the organic pollutant molecules/^•OH radicals contact, and accelerating the degradation. According to Deyou et al. [203], the particle size and surface area are more important factors than a crystalline structure for improving the catalytic efficiency of CuFe₂O₄. López-Ramón et al. [204] evaluated the effect of calcination temperature on the catalytic activity of CuFe₂O₄ synthesized by the sol–gel method, finding that the calcination temperature has two opposite effects: the activity decreases with increasing temperature due to the increase in crystalline size and cubic-to-tetragonal transformation of ferrite and appearance of hematite; however, the metal leaching decreases with increasing calcination temperatures.

Cobalt ferrite (CoFe₂O₄), like other ferrites, has a high catalytic activity, stable crystal structure, low solubility, magnetic properties, as well as the ease of controlling the leaching of cobalt due to the strong interactions between metals, and the strong redox activity of Co with catalytic properties similar to those of noble metals (Pt, Ir, and Au) [205,206,207,208], which make it a promising material as a Fenton-like catalyst.

Feng et al. [209] synthesized monodispersed CoFe₂O₄ nanoparticles by a solvothermal method to evaluate them as a Fenton catalyst in the degradation of methylene blue, reaching a concentration decrease of 96.8%. The authors highlighted that the existence of Co²⁺ ions could favor the decomposition of H₂O₂ to OH^• and subsequently give way to different reactions, as evidenced in Equations (11)–(14).

$${\mathrm{C}\mathrm{o}}^{2+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{C}\mathrm{o}}^{3+}+{\mathrm{O}\mathrm{H}}^{-}+{\mathrm{O}\mathrm{H}}^{•}$$

(11)

$${\mathrm{O}\mathrm{H}}^{•}+{\mathrm{H}}_2{\mathrm{O}}_2\to \mathrm{O}{\mathrm{O}\mathrm{H}}^{•}+{\mathrm{H}}_2{\mathrm{O}}_2$$

(12)

$${\mathrm{F}\mathrm{e}}^{3+}+\mathrm{O}{\mathrm{O}\mathrm{H}}^{•}\to {\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{H}}^{+}+{\mathrm{O}}_2$$

(13)

$${\mathrm{C}\mathrm{o}}^{3+}+\mathrm{O}{\mathrm{O}\mathrm{H}}^{•}\to {\mathrm{C}\mathrm{o}}^{2+}+{\mathrm{H}}^{+}+{\mathrm{O}}_2$$

(14)

Sing and Singhal [210] demonstrated that the transition-metal doping of cobalt ferrites is a promising method for tuning the physical characteristics of catalysts and thus, enhancing their catalytic and magnetic properties. For that, the authors synthesized a series of Ru-doped cobalt ferrite nanoparticles by the sol–gel method for the photo-Fenton degradation of red Remazol textile dye, achieving a degradation of approximately 90% within 120 min. The mechanism proposed is based on the photocatalytic and Fenton character of the Ru-modified ferrite. An electron-hole pair is created by the irradiation of cobalt ferrite nanoparticles with visible light. The photogenerated electrons are responsible for the OH^• generation from H₂O₂ and also the reduction of the Fe³⁺ cation on cobalt ferrite to Fe²⁺, which further generates OH^• radicals in the reaction with hydrogen peroxide. Vinosha et al. [211] also analyzed the photo-Fenton performance of CoFe₂O₄ nanoparticles obtained by means of chemical coprecipitation, achieving almost total degradation of methylene blue (~99.3% in 75 min) under visible light irradiation. As an outstanding result, they proposed that the pH used in the synthesis was not an impact parameter that affected the morphology of the catalyst; however, it significantly affects the particle size (a more alkaline (pH > 9) medium, larger crystallite size). It has been proposed that the reactions that lead to the formation of CoFe₂O₄ by the chemical coprecipitation method in an aqueous medium, are those presented in Equations (15) and (16) [212].

$${2\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{C}\mathrm{o}}^{2+}\to {\mathrm{F}\mathrm{e}}_2\mathrm{C}\mathrm{o}({\mathrm{O}\mathrm{H})}_8$$

(15)

$${\mathrm{F}\mathrm{e}}_2\mathrm{C}\mathrm{o}({\mathrm{O}\mathrm{H})}_8\to \mathrm{C}\mathrm{o}{\mathrm{F}\mathrm{e}}_2{\mathrm{O}}_4+4{\mathrm{H}}_2\mathrm{O}$$

(16)

In turn, Iron cobaltite (FeCo₂O₄) has been also studied in environmental remediation and energy storage, thanks to its electrical properties and electrochemical performance [213]. In the energy storage field, Mohamed et al. demonstrated that iron cobaltite nanorods show a better capacity and lower overpotential as the cathode of lithium−O₂ batteries than other metal cobaltites (Mn, Ni, and Zn) [214] because the FeCo₂O₄ surface has the highest number of electropositive Co³⁺ active sites that improve the oxygen adsorption and Fe²⁺ in the tetrahedral site that favors the release of electrons to reduce oxygen. Yadav et al. [215] demonstrated that iron cobaltites are also efficient for supercapacitive and photocatalytic applications due to the valence states of the Fe³⁺/Fe²⁺ and Co³⁺/Co²⁺ species, which are considered active catalytic sites. These nanoflake-like iron cobaltites present a capacitance as high as 1230 F g⁻¹ (5 mV s⁻¹) with a good rate capability and superior cycling stability and also show a good photocatalytic performance achieving up to 94.19% degradation of crystal violet dye under sunlight illumination. However, despite this, very little work has been carried out with reference to their evaluation as a Fenton-like catalyst [216,217]. Zhang et al. [216] synthesized nitrogen-containing carbon/FeCo₂O₄ composites and analyzed their performances as Fenton catalysts for the degradation of methylene blue obtaining almost 100% removal in 10 min without pH adjustment, which was attributed to the uniform distribution of bimetals and nitrogen doping, which ensured the exposure of sites with high catalytic activity. Zhao et al. [217] analyzed the behavior of FeCo₂O₄/g–C₃N₄ as a photo-Fenton catalyst in the degradation of rhodamine B (RhB), obtaining 98% degradation in 45 min, which was attributed to a synergetic interaction between photocatalytic and Fenton-like reactions and the effective separation of the photogenerated charges. Therefore, it is proposed that the use of iron cobaltite as a pristine catalyst (without doping or support) in the Fenton reaction may be promising for future applications.

In general, spinel ferrites turn out to be attractive materials for catalytic activities in Fenton processes, mainly because they manage to improve the generation reaction of Fe²⁺ and achieve a synergistic effect between metal ions with valences similar to those of iron involved in the process (M²⁺, M³⁺, M: Metal).

Perovskites are other types of materials with promising catalytic activity that can be synthesized by the abovementioned methods (Chemical coprecipitation, sol–gel, and hydro/solvothermal) [218,219,220,221]. Perovskites can be defined as a type of mixed oxide with different formulations, binary (ABO), ternary (AA’BO or ABB’O), and quaternary (AA’BB’O), where A and B are cations sites occupied by alkali metals, alkaline-earth metals or rare-earth metals and transition metals, respectively [222].

Some perovskites have been studied in different Fenton-type reactions, for example, Carrasco-Díaz et al. [75] removed paracetamol by Fenton reaction using LaCu_1−xM_xO₃ (M = Mn, Ti) as the catalyst and determined that the most active catalyst was the one that contained the highest amount of Cu²⁺ at the surface. Moreover, they found that the titanium and manganese species seem not to be responsible for the improvement of activity with respect to the sample LaCuO₃. Li et al. [223] synthetized a Ca_1−xFeO_3−δ perovskite and determined that the A-site cation can distort the FeO₆ octahedra in the perovskite and regulate the oxygen vacancies (OV) concentration; in this way, an A-site cation deficient of Ca_0.9FeO_3-δ results in an improved H₂O₂ activation for the degradation of tetracycline by a Fenton-like process. Similarly, Xie et al. [224] found that the copper incorporation in LaCoO₃ perovskite improved the electro-Fenton activity due to the enhancement of redox activity and oxygen vacancies, but in this case, by the substitution of B-site elements, which synergistically promoted the activation of hydrogen peroxide to a hydroxyl radical (^•OH). On the other hand, Rusevova et al. [225] degraded phenol via heterogeneous Fenton-like reactions using iron-containing LaFeO_3, and BiFeO₃ perovskites, and made a comparison with data reported in the literature using, as a catalyst, nano-sized Fe(II, III) oxide particles, determining that the perovskites synthetized in this work had a higher catalytic activity. Zhao et al. [226] determined that BiFeO₃ supported in carbon aerogel (BFO/CA) with a three-dimensional (3D) structure improves the catalytic activity and stability of BiFeO₃, resulting in an interesting strategy for the development of advanced catalysts for its possible application in Fenton processes.

In summary, perovskites are materials similar to spinel ferrites, and the contribution of two different metal species (formation of OVs) can be of interest to the Fenton process.

5. Oxygen Reduction Reaction through Two Electrons (ORR 2e⁻)

The oxygen reduction reaction (ORR) has attracted attention in recent years due to its applicability in various fields of industry, both in energy storage and water treatment processes [227,228,229]. However, this review is focused on ORR as a route for the electrogeneration of H₂O₂, a product that controls the efficiency of the electro-Fenton oxidation reaction.

In general, ORR involves two types of reactions [230]:

• Oxygen molecules that are converted directly into water (4e⁻);
• Oxygen molecules that are initially reduced to peroxide and later to water (2e⁻).

Analyzing the mechanism, it has been established that ORR 2e⁻ takes place in two steps involving two single proton-coupled electron transfers; in the first, there is the transfer of electrons and generation of the HOO intermediate linked to the active site (*), while in the second, H₂O₂ is produced (Equations (17 and (18)).

$${\mathrm{O}}_2+\mathrm{*}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\to \mathrm{O}\mathrm{O}{\mathrm{H}}^{\mathrm{*}}$$

(17)

$$\mathrm{O}\mathrm{O}{\mathrm{H}}^{\mathrm{*}}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\to {\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{*}$$

(18)

However, H₂O₂ types formed at the cathode can react in three ways [231]:

• Being electro-reduced to H₂O at the cathode (usually at porous cathodes);
• Being oxidized to O₂ at the anode via HO₂· as an intermediate;
• Disproportionate to O₂ and H₂O in a non-electrochemical reaction.

Therefore, the objective in the design of catalysts for the electroproduction of H₂O₂ by ORR is to synthesize materials capable of binding the OOH* intermediate with the appropriate strength; this is neither very strong nor weak to avoid the 4e⁻ pathway predominating [232].

The most studied way of improving the performance of the ORR 2e⁻ is the synthesis and optimization of the cathode materials, since, together with the generation of H₂O₂, the current efficiency and, therefore, the energy consumption needs can also be taken into account [233]. In general, an optimal catalyst for ORR 2e⁻ must have good electrical conductivity, stability, and resistance to acid corrosion [234], properties associated with carbon-based materials [235]. Carbon materials have proven to have excellent properties for use as a cathode due to their high electrical conductivity, zero toxicity, low cost, high stability, and high catalytic activity [236,237,238]. However, although carbonaceous materials exhibit high selectivity toward H₂O₂ production, they have a slow kinetic toward this reaction, resulting in poor performance.

Different strategies to improve the electrochemical production of H₂O₂ have been recently studied and one of them consists of the deposition of other catalysts or metal oxides (MnO₂, V₂O₅, and CeO₂) on the surface of carbonaceous materials [155]. On the other hand, doping with heteroatoms such as oxygen (O), sulfur (S), nitrogen (N), and fluorine (F) is another strategy developed to increase the electrocatalytic performance of carbon materials. The carbon lattice doping with heteroatoms introduces defects to the carbon structure, modifying their atomic and electronic structure and, thus, resulting in new promising materials for several energetic applications such as energy production and storage. However, the improvement mechanism of the introduction of these heteroatoms is still under debate by researchers worldwide. Some authors [239,240,241,242,243] found that the introduction of some oxygen functional groups can induce greater electrical conductivity and electrocatalytic activity thanks to the formation of a hydrophilic surface. Recently, different carbonaceous materials have been evaluated as ORR 2e⁻ catalysts, where the importance of the amount and type of oxygenated functional groups present on the surface of these has been evidenced, finding that a greater amount of carboxyl group translates into greater efficiency in H₂O₂generation because it facilitates the displacement of the electron density of the active site, allowing a better interaction in the adsorption of O₂ [162,244]. In the same way, it has also been established that the higher the ratio of I_D/I_G in carbonaceous materials, the greater the production of H₂O₂ [245].

Zahoor et al. [246] indicated that the sp² electron configuration of graphite, graphene, and nanotubes are rich in π electrons which can be transferred without resistance in ORR; however, this electron flow is not enough to promote ORR. Thus, the heteroatom introduction (e.g., N) within its network induces π electron activation by bonding with the solitary electron pairs of N and providing negative charges. The carbon atoms adjacent to nitrogen have a high positive charge density, which increases the adsorption of oxygen and reactive intermediates, which results in enhanced ORR activity. Okamoto et al. [247] found by density-functional calculations that the adsorption of O₂ is more energetically favorable with the increase in the number of N around a C=C bond. Lagarreta-Mendoza et al. [248] proposed quantum phenomena for the explication of the ORR mechanism on N-doped carbon-based electrocatalysts, which consists of the hybridization change in carbon atoms in the graphene lattice, from sp² to sp³. Nitrogen doping could induce a quantum tunnel phenomenon called nitrogen inversion, which consists of a process in which the lone pair of an sp³ nitrogen atom migrates from one face of the atom, travels through the nucleus (quantum tunneling), and reappears on the other side. This inversion of nitrogen (due to network defects) allows the N atom to act as a “switch” that activates or stops the flow of electrons, generating active sites for ORR The heteroatom–carbon interaction induces the activation of the electron π when joining with the lone pairs of N so that later, the O₂ molecules are reduced into the carbons positively charged by the neighboring N. Additionally, it has been observed that the insertion of a single atom of N does not, apparently, have enough influence to alter the carbon, so a minimum of three atoms are necessary to cause small differences in the length of the C and N bonds [248]. As reported in the literature, N-pyrrolic is more likely to promote ORR 4e⁻ by increasing electron donation capacity, while N-pyridinic has more tendency for ORR 2e⁻ [249]. Wang et al. [250] investigated the ORR 2e⁻ activity of multiwalled carbon nanotubes (MWCNT) with different degrees of oxidation by oxidation with concentrated sulfate and potassium permanganate at 20 to 60 °C for 1 h. They observed that the presence of oxygen groups created after the oxidation treatment enhances the H₂O₂ generation, with the MWCNT always being oxidized and more active than the nonoxidized counterpart. Among the oxidized MWCNT, the sample prepared at middle conditions (40 °C and 1 h) present the best H₂O₂ selectivity which was explained by the composition of its outer and inner structure. After the oxidation treatment, the outer surface is damaged by the creation of a large number of defects together with the introduction of oxygen functional groups such as COOH and C-O-C that act as excellent active sites for the ORR 2e⁻, while the inner structure is maintained, ensuring its good electrical conductivity.

Similarly, it is reported that carbon doping with heteroatoms and transition metals can energetically optimize the charge and rotation density of carbon, helping to promote the activation and reduction of O₂ by weakening the O–O bond in the ORR process [251]. Some studies have found that N can change the mode of adsorption of O₂, while sulfur (S) affects the binding capacity of the OOH intermediate by varying the spin density [252,253]. Ting Soo et al. [254] reviewed the electrochemical performance of modified graphene as an ORR electrocatalyst, and the spin density, charge, and energy gap were evaluated, finding that the microstructural properties of heteroatom-doped graphene and the high spin density or atomic charge density on adjacent carbon atoms are the key factors that contribute to the enhanced electrocatalytic activity. However, an over-dose of heteroatom doping has negative effects due to the formation of high-sized clusters that make electron transfer difficult. Zhang et al. [255] explained that any chemical species that is substituted for graphene can lead to a high asymmetry spin density and atomic charge density, which promotes ORR. However, by applying DFT calculations they demonstrated that ORR on N-doped graphene is a four-electron pathway. Conversely, Pei Su et al. [256] prepared N-doped graphene showing high H₂O₂ yield, selectivity, and low EEC. They also observed that the C=C and N-pyridinic bonds in N-graphene have the ability to enhance the onset potential, while the N-graphitic groups are responsible for the current disk obtained. Moreover, in EF reaction, they demonstrated that N-graphitic are responsible for the H₂O₂ generation, and N-pyridinic groups catalyze the OH^• radical productions from H₂O₂ (Figure 10). To further investigate the effect of the N-group amount and nature, Pei Su et al. [257] controlled the N-surface chemistry of N-doped graphene (N–GE) via pyrolysis temperatures (200, 300, 400, 550, 750, and 950 °C) in the catalytic activity, observing that the value of electrons transferred in the ORR was in the range of 2 to 2.5 for temperatures of 200 to 400 °C, and from 2.5 to 2.7 for further temperature increases. Thus, the sample pyrolyzed at 400 °C presented the highest active N content and H₂O₂ selectivity and achieved the highest phenol degradation (93.58%). They proposed that the conversion of graphite N and pyridinic-N in N–GE plays an important role in the oxygen reduction reaction (ORR) and OH^• conversion, while the conversion of pyridinic-N-oxide to pyridinic-N is critical for catalyst stability and sustainability.

Qin et al. [258] and Xie et al. [259] evaluated N and O co-doped carbon material, highlighting the synergistic effect of the different species composed of the two atoms in question on the generation of H₂O₂. The pyridine-N groups combined with ether (C-O-C)/carboxyl (COOH) were highlighted, which presented lower adsorption energy of the intermediate *OOH, facilitating protonation toward H₂O₂. Some studies [260] have also pointed out that the -COOH and carbonyl (-C=O) groups are active for ORR 2e⁻, and mainly the -COOH group.

Iglesias et al. [261] synthesized N-doped graphitized carbon nanohorns (N–CNH) via coating and controlled annealing of polydopamine (PDA) (Figure 11) and analyzed its use as an electrode material for OEE 2e⁻. The excellent catalytic performance of N–CNH to perform ORR 2e⁻ is explained based on the high surface area and accessible porosity of the CNH scaffold, a good distribution of different N-contained surface groups, and improved conductivity and easy electron transfer. The authors also highlight that the microporosity and the specific pyridinic/pyrrolic ratio are crucial factors affecting the ORR route. The microporosity decreases the residence time of H₂O₂ on the material avoiding further reduction to H₂O₂. The relative types of N atoms also affect the electrocatalytic performance. However, the influence is dependent on the pH. At an acidic pH, protonation of the N-pyridinic occurs which compromises the ability to weaken the O–O bond producing H₂O₂ as the main product while at a neutral or alkaline pH, a reduction in selectivity toward H₂O₂ is observed because the formation of H₂O by the O–O breaking becomes progressively more available.

Chen et al. [262] also demonstrated that porosity is so important in controlling the selectivity of the process. Chen et al. determined that activated carbon with abundant micropores and a high surface area is better for high ORR activity via 4e⁻, while carbon and graphite black with a smaller specific area tend to catalyze ORR via 2e⁻. Granules with more micropores facilitated the ORR process and caused a further reduction of H₂O₂, leading to less H₂O₂ generation. Similarly, Wenhui He et al. [263] analyzed the effect of the 3D hierarchical porosity on N-doped carbon on the catalytic performance. For that, they prepared N-doped carbon with hierarchical micro-, meso-, and macroporosity by the use of 3D macroporous templates and endogenous pore-generating agents with the assistance of KOH activation. It was concluded that the catalytic efficiency of nitrogen-doped species in ORR highly depends on their degree of exposure to the carbon matrix and that the hierarchical structure can strongly influence the rate of diffusion of O₂ and rapid release of H₂O₂.

In general, it can be concluded that there are some important requirements to obtain carbonaceous materials with high catalytic activity for ORR 2e⁻ such as high porosity (subject to further investigation), large surface area, presence of oxygenated groups, bonds C=C, N-pyridinic, and a greater amount of C atoms sp³ hybridization.

Zhang et al. [264] synthesized interface atomic domains (IAD) of C/N/O atoms (pyridine-N, C=O), from biomass tar by pyrolysis, oxidation, and doping. With this assured an ordered structure (by pyrolysis), with functional groups C=O (oxidation), which results as a host for the generation of N-pyridinic (doping). As a result, it was found that the material obtained had high selectivity for ORR 2e⁻ (N-pyridinic) and an excellent mass transfer of O₂ (C=O and ordered structure).

Another strategy is the doping of porous carbon materials with transition metals which can accelerate the redox reaction by the action of metal ions, thereby increasing the rate of electron transfer. Similarly, nonprecious metals have proven to have an excellent capacity for ORR [265]. It has been shown that compared to bare carbon, metal/carbon compounds have a noticeable improvement toward ORR [266]. Metal/carbon catalysts can adsorb O₂ in two ways, which translates into a greater or lesser efficiency toward reduction via 2 or 4 electrons [267,268,269,270].

• Lateral adsorption leads to a longer bond length and therefore weakens the O=O bond, giving as a product, H₂O (Figure 12).
• Adsorption at the end, where oxygen is adsorbed in the form of OOH, which can produce H₂O₂ and H₂O (Figure 12).

Thus, to obtain metal-doped materials selective to the electrogeneration of H₂O₂, the end-on adsorption of O₂ molecules on the catalyst surfaces need to be promoted. One strategy is the isolation of the active atomic metal sites [271]. For that, three pathways have been developed: (i) alloying an active metal with an inert metal that can induce geometric isolation of the active metal where the O₂ is end-on adsorbed; (ii) the partial poisoning of the active metal surface with an inert phase or molecules, e.g., the coating of Pt nanoparticles with amorphous carbon by CVD (Figure 13); and (iii) the synthesis of atomically dispersed catalysts.

Some nonprecious metals, used as a doping material in carbonaceous matrices, despite increasing costs, are offset by high dispersion in the matrix, which favors the production of H₂O₂ [273]. In general, although iron may be an ideal metal to regulate the binding energy of the intermediate *OOH with the carbon matrix, it has been shown that it has a strong tendency toward the route of 4e⁻ [274,275]. Co is one of the metals studied for ORR, demonstrating greater selectivity toward two electrons compared to Fe [276]. On the other hand, Chen et al. [277] investigated the incorporation of Cu atoms in mesoporous N/O-codoped graphitic carbon catalysts with a tailored mesoporosity concluding that the H₂O₂ production depends on the number of C atoms bound to the O atom, which increases by doping with a tiny amount of N. However, the inclusion of copper significantly decreased the selectivity toward H₂O₂. Cu, having a strong binding with oxygen, quickly generates blockages of the active sites due to the O and OH species generated in the reaction [278]. In other studies, Zhang et al. [279] synthesized Au@Cu_2−xS-CNTs core-shell deficient in Cu and determined that the catalysts exhibited higher ORR activity and selectivity toward H₂O₂, associated with the Cu defects, which (from theoretical calculations) decreases the energy of reaction of OOH*. Wang et al. [280] probed that the nanostructure of the Cu active site is the key to the success of the catalyst; if the active site and host effect are optimally coordinated, it will certainly show considerable potential in the catalytic performance. For that, the authors designed electron-deficient and electron-rich Cu active sites by the regulation of nitrogenous ligand solvents and analyzed the effect in the capture and activation of the C≡C bond of acetylene. The results showed that the Cu sites, stabilized by N-pyrrolic, generate electron-deficient states that improve the activation capacity of acetylene and, thus, the catalytic activity (hydrochlorination). In turn, Hyo-Noh et al. [281] determined that Cu can strongly increase the chemical bonding with O (unfavorable for ORR 2e⁻) and thus bare Cu particles cannot catalyze the ORR at all, but this energy can be easily controlled by the coating with N-doped carbon shells (regulated through CO₂ treatment), resulting in ORR activity comparable to the Pt (111) surface.

Therefore, recently, the investigate into other types of metals has begun. Lenarda et al. [282] investigated different metal (Fe, Ni, Mn, and Co) nanoparticles embedded with an N-doped graphitic carbon and determined that the nature of the metal dictates the type of final surface N-group by obtaining different N-pyridinic/N-pyrrolic ratios, which is essential for H₂O₂ selectivity. The optimal ratio was obtained when Co was used.

Thus, Co is presented as a promising metal to increase the ORR activity without losing H₂O₂ selectivity. In a recent study, Gao et al. [283] prepared single-atom transition metals anchored in nitrogen-doped graphene citing that Co could be an active site for the ORR 2e⁻ pathway since the bond with the OOH* intermediate has optimized adsorption energy, neither too strong nor too weak, in addition to presenting a higher OOH* to O* reduction barrier than that achieved with iron and magnesium atoms. However, Cao et al. [284] evaluated Co nanoparticles supported on N-doped holey graphene aerogel in ORR, determining that the reaction pathway was given by a 4-electron process. Collman et al. [285] studied the geometric effect of cobalt on two different porphyrin structures (monomeric and dicobalt), where the monomeric consists of a Co atom surrounded by N, with which they were able to determine that the selectivity toward H₂O₂ is favored mainly by the isolated site above the electronic effects, since in both cases, the union of the OOH* is the same. In turn, Ferrara et al. [286] studied the effect of pyrolysis temperature against the selectivity of H₂O₂ in Co-encapsulated nanoparticles on a nitrogen-doped carbon matrix, determining that higher temperatures (>800 °C) lead to greater H₂O₂ selectivity, attributed to a larger degree of graphitization and pyrrolic N content. However, the N and O content progressively decreases with increasing pyrolysis temperature, leading to a lower ORR current density.

However, some of the problems of catalysts doped with nonprecious metals are related to their instability in selectivity toward H₂O₂ with pH. At acidic pHs, the 2e⁻ route is favored; however, at basic pHs, the 4e⁻ route is given. Therefore, the use of metal oxides such as magnetite has been proposed to improve the ORR 2e⁻ activity of catalysts based on carbonaceous materials [287].

With the above, and considering the other limitations associated with both the ORR 2e⁻ and Fenton reactions, the doping of carbonaceous materials with structures of the spinel ferrite type is proposed as a viable solution, where the desired bifunctionality is achieved, both for obtaining H₂O₂ and its subsequent decomposition (Fenton) to ^•OH. This fact, together with the metal active site control and the spontaneous radical ^•OH production in controlled N-doped carbons, opens the gate to the design of new bifunctional catalysts for the simultaneous production of H₂O₂ and its conversion to ^•OH radicals for the electro-Fenton oxidation of organic pollutants in the water.

6. Bifunctional Electro-Fenton Catalyst for Direct ^•OH Formation

Despite the advances achieved in EF and ORR for the on-site production of H₂O₂, there are still some problems to overcome such as the secondary pollution related to the Fenton catalysts [288,289] and the requirement of two catalysts: one selective to the reduction of oxygen to H₂O₂ and another Fenton-type catalyst for the transformation of H₂O₂ to hydroxyl radicals. Recent efforts are being made to develop materials with dual functionality for the electroreduction of oxygen to H₂O₂ and Fenton catalysis. However, the preparation of heterogeneous EF catalysts with high selectivity and activity toward ORR via the two-electron pathway is challenging since the active sites for Fenton catalysis are mainly transition metals that usually catalyze oxygen reduction via the 4e⁻ route without the production of H₂O₂.

One strategy to develop bifunctional catalysts capable of generating and activating H₂O₂ is the doping of carbon materials with different heteroatoms or carbon structures. Li et al. [290] demonstrated that heteroatom-doped carbon materials can convert H₂O₂ to ^•OH radicals without the need for a Fenton metal catalyst. For that, the authors synthesized an O, F-codoped carbon bifunctional catalyst by the carbonization of polyvinylidene fluoride and proposed that the H₂O₂ generation depends principally on the ratio C-O/C=O (optimal value in 4), while the activation to ^•OH occurs on the semi-ionic C-F bonds obtaining a sulfamerazine degradation of 90.1% in 3 h during the metal-free EF process. A similar fact was observed by Yang et al. [291] using N-doped graphene. They found that the bifunctional effect is attributed to the presence of N-graphite, which influences the H₂O₂ generation and N-pyrrolic responsible for H₂O₂ activation. Similarly, Haider et al. [292] synthesized, in situ, an electrode from polyaniline-derived N-doped carbon nanofibers with bifunctional activity in EF, which was attributed to the content of C=C, oxygen groups, and N-graphitic, which enhanced the H₂O₂ generation and which can be further activated by N-pyridinic to generate the ^•OH radicals. In turn, Qin et al. [293] used it as a bifunctional metal-free cathode based on O-doped carbon nanotubes. They attributed the excellent metal-free EF behavior of this catalyst to the defects and the C-sp³ that enhances the oxygen adsorption promoting the H₂O₂ production and to the -C=O active sites that are associated with the ^•OH production. Chen et al. [294] found that groups BC₃ and BC₂O present in a boride-activated carbon enhanced the oxygen adsorption and facilitated the desorption of ^•OOH, improving the H₂O₂ generation but also promoting the in situ conversion to ^•OH. Xie et al. [295] showed that the reduced graphene oxide could improve the H₂O₂ generation, while the MOF (MIL-88A) and carbon felt (MIL-88A/CF) could activate the H₂O₂ to OH^•; thus the cathode rGO/MIL-88A/CF resulted in an excellent bifunctional catalyst.

The introduction of different metals in the carbon matrix for use as bifunctional catalysts has also been studied from simple immobilization to more complex structures such as core-shell. Zhang et al. [296] immobilized Fe₃O₄ in a gas diffusion electrode (GDE) and used it as a rotating cathode for the EF removal of tetracycline. They concluded that the rotating disk gas diffuser greatly enhanced the mass transfer providing a high efficiency in the H₂O₂ generation, whereas the Fe₃O₄ active phase was capable of dissociating the H₂O₂ to OH. Cui et al. [297] also developed a Fe₃O₄/MWCNTs/CB (carbon black) GDE, which has high performance in the production and activation of H₂O₂. The presence of CB enhanced the H₂O₂ production, whereas its activation to OH^• was ascribed to ≡Fe(II) on the surface of Fe₃O₄/MWCNTs. Moreover, an improvement in the regeneration of Fe²⁺ was observed due to the high-speed charge channel of MWCNT (Figure 14). On the other hand, Hu et al. [298] synthesized a bifunctional catalyst based on Fe₃C and FeN nanoparticles encapsulated by porous graphitic layers and determined by density function calculations that the active sites for H₂O₂ generation were the Fe₃C, and the activation to OH^• depend on FeN_x sites. Cao et al. [299] designed and synthesized FeO_x nanoparticles embedded into N-doped hierarchically porous carbon for its use as a bifunctional catalyst against phenol, sulfamethoxazole, atrazine, rhodamine B, and 2,4-dichlorophenol in neutral water solution. They found that the presence of N pyridinc and N-pyrrolic increased the H₂O₂ selectivity, and the FeO_x nanoparticles could enhance the electron transfer to improve the OH^• production. Moreover, the Fe^3+/Fe²⁺ cycle could be enhanced by the strong interaction between N-doped porous carbon and FeO_x NPs.

Another proposal to improve the production and activation of H₂O₂ has been the use of catalysts with the presence of two or more different metal or metallic species that can moderate the activity of alone metals and thereby obtain better bifunctional activity. Ghasemi et al. [300] prepared a cathode with carbon nanotubes (CNTs) and CuFe nano-layered double hydroxide showing that the incorporation of CNTs improves the H₂O₂ generation, and the atoms of Cu and Fe were responsible for the generation of OH^•. Cui et al. [301] used Cu/CuFe₂O₄ integrated graphite felt as a bifunctional cathode and attributed the H₂O₂ generation to Cu⁰ and the OH^• production to primary Fe²⁺ ions. Moreover, the Fe³⁺/Fe²⁺ redox cycles could be accelerated by the electron donation of Cu⁰ and Fe⁰, and even Cu⁺. Similarly, Luo et al. [302] degraded 98.1% of tetracycline in 2 h using Cu-doped Fe@Fe₂O₃ core-shell nanoparticles loaded on nickel foam as a cathode, and determined that Cu⁰ and Fe⁰ react with O₂ producing H₂O₂; Fe²⁺ and Cu⁺ are responsible for the ^•OH generation from the in situ generated H₂O₂. Moreover, they proposed that Fe²⁺ and Cu²⁺ additional EF active sites could be generated from the redox process between Cu⁺ and Fe³⁺ and the replacement reaction between Cu²⁺ and Fe⁰. Sun et al. [303] synthesized a nickel foam cathode co-modified with core-shell CoFe alloy/N-doped carbon (CoFe@NC) and carbon nanotubes (CNTs), where the production of H₂O₂ was attributed to CNTs and the activation to OH^• to CoFe@NC through Fe²⁺ and Co²⁺ oxidation to Fe³⁺ and Co³⁺, which are again reduced by Co⁰/Fe⁰ and the electric field. Similarly, Yu et al. [304] determined that the carbon felt was the ORR catalyst selective to H₂O₂, and the Co and Fe atoms generated the OH (Co > Fe) in the cathode of CoFe-layered double hydroxide/carbon felt with bifunctional activity. Li et al. [305] prepared an electrode doped with Fe and Mn oxides coated with a carbon layer and determined that the carbon material was responsible for H₂O₂ generation, and the metals for OH^• production. It was also noted that the presence of Mn considerably improves the regeneration of Fe³⁺/Fe²⁺ due to Mn that could transfer an electron to Fe to accelerate the cycle.

Bifunctional catalysts using a 3-electron ORR pathway have also been recently proposed as an improved alternative. Xiao et al. [306] reported a catalyst capable of directly generating ^•OH via a 3-electron pathway with FeCo alloy encapsulated by carbon aerogel. This performance was principally associated with the generation of H₂O₂ by the 2 electron ORR of the COOH group present in the graphite shell, followed by the reduction by one electron toward ^•OH. This 3-electron pathway can be regulated by the electrons from the encapsulated metals.

7. Conclusions

This paper summarizes the different strategies for the design of catalysts with high performance toward Fenton processes and the ORR by the 2e⁻ route as a mechanism to improve the efficiency of the electro-Fenton process, finally focusing on the design of bifunctional electro-Fenton catalysts for the production and activation of H₂O₂.

The key to obtaining high efficiency of ORR 2e⁻ is the proper use of structures that allow the end-on adsorption of O₂, oxygenated functional groups on the surface of the catalyst, as well as the inclusion of doping atoms that allow synergistic effects toward the selectivity of H₂O₂. Although the strategies developed have been based on the amount of H₂O₂ produced and the ORR pathway, some research has opened a gate to the possibility that the morphology of the surface of the catalyst has a much greater impact than expected. Therefore, it is necessary to evaluate, in detail, not only the chemical composition but also other factors such as hydrophobicity, roughness, and/or porosity of the catalyst surface. In general, there are several parameters to consider in the search for an ideal catalyst for ORR 2e⁻, but the amount of N-pyridinic, hydrophobicity, roughness, and doping metal, are so far the most crucial.

As Fenton catalysts, different Fenton-type metals and bimetallic-based materials have been reviewed, raising spinel ferrites as an alternative of interest to provide the bifunctionality for carbonaceous materials, by improving selectivity toward ORR 2e⁻, maintaining the metal ions necessary for Fenton-type catalytic activity suitable for generating OH^• obtained from the H₂O₂ generated in the same structure of the catalyst.

Finally, bifunctional electro-Fenton catalysts are described for the production and activation of H₂O₂. The main strategies are focused on the use of heteroatom-doped carbon materials or carbon-encapsulated Fenton-type metals/carbon composites that moderate the metal’s activity and provide an active site for H₂O₂ generation by the ORR 2e⁻ route, presenting metal sites for the activation of in situ generated H₂O₂ to ·OH radicals required for the degradation of organic pollutants.