Considering WoS and Scopus databases, the search identified a total of 1073 and 1653 articles in indexed journals, respectively, until 2022, with the earliest indexed publication on the theme in Scopus dating back to 1992. Among these articles, 141 and 390 contained the term “Fenton” in their title but these did not employ the technique in their study. These papers were excluded from further analysis as they were not aligned with the objectives of this systematic review. In total, 932 and 1263 articles (WoS and Scopus databases, respectively) on the implementation of the Fenton technique to remove any form of pollutant were considered.

In regard to the use of the WoS database, it is important to indicate that WoS is mainly designed for researchers to find published literature, and it could be restricted for bibliometric studies. Then, it is not suitable to directly use all different types and levels of the WoS database for bibliometric study. For example, Data Citation Index, Derwent Innovations Index, Zoological Record, Social Sciences Citation Index, Arts & Humanities Citation Index, Conference Proceedings Citation Index—Social Sciences & Humanities, Book Citation Index—Social Sciences & Humanities, Current Chemical Reactions, and Index Chemicus. In fact, a minor number of articles was identified by the WoS search. Then, the information was collected once again with the same methodology previously used but the Scopus database was utilized.

Comparing the results obtained in the two databases (Figure 1), it is observed that WoS is less comprehensive than Scopus. This may be due to the Scopus database having a larger, internationally diverse database as well as a range of intelligent tools that can be used to aid the research process, such as journal rankings, author profiles, the number of articles published by a journal in a given year, and the frequency of use of scientific terms. Therefore, Scopus was used to elaborate the systematic review, examining the results obtained. Additionally, WoS results were reported as Supplementary Materials (SM) in order to illustrate the similarities.

Figure 1 illustrates the increase in the number of articles over time (Pearson’s r = 0.95 and 0.93, for WoS and Scopus databases, respectively, with p < 0.001); a more pronounced growth in publications is attained since 2013 and, notably, in the years 2020 and 2022.

The findings revealed a trend over a period of thirty years in which the number of publications featuring AOPs has increased. This pattern is consistent with the macro-level explorations of chemical, electrochemical, photochemical, and photoelectrochemical technologies designed to remove organic pollutants from wastewaters [18]. The growing demand for the treatment of effluents that contain complex and recalcitrant pollutants is driving the use of the Fenton method [19]. Furthermore, the Fenton process is a versatile technique that can be combined with other technologies, as highlighted by Ramos et al. (2021) [20], who described the use of Fenton in a systematic review on textile dye treatment. Fenton is a homogeneous reaction, which occurs via a chemical reaction of the mixture of Fe²⁺ and H₂O₂, which leads to the rise of homogeneous hydroxyl radicals (^•OH) [9]. The use of Fenton is due to the growing demand for wastewater treatment containing increasingly complex and difficult to remove recalcitrant pollutants. It means that, with this type of oxidant, it is possible to treat different and varied ranges of residues, allowing for a low cost and reaction time, as well as avoiding the use of large technological devices [9].

From 2017 to 2022, researchers have incorporated statistical tools such as Takeuchi’s method using the Mini Tap program, as well as the response surface methodology (RSM) with designs including the Box–Behnken design, Box–Wilson statistical design, central composite design (CCD), Central Composite Design in MINITAB^®16 software, and other arrangements, such as the orthogonal array test and single-factor experiments, orthogonal experiments, and the regression quadratic equation [21,22,23,24,25,26,27,28]. Furthermore, the use of new reaction forms has increased, as described below.

Table 1 displays the nine categories of pollutants that were degraded using Fenton during 1992–2022, across 1263 studies. Among these studies, 90 focused on the use of Fenton to concurrently degrade various compounds, specifically those falling under the categories of pharmaceuticals, hygiene/cleaning products, and dyes.

The table above displays the results of studies that focused on the removal of pollutants using the Fenton technique. Among these studies, the majority focused on the removal of dyes (227; 18%), organic compounds (216; 17%), and pharmaceuticals (114; 9%); fewer were focused on removing phenol (62; 5%), inorganic compounds/metals (40; 3%), pesticides (22; 1.7%), hygiene and cleaning products (11; 0.87%), and vinasse (5; 0.39%). However, a large proportion of the studies (418; 33.1%) examined the use of the Fenton technique to remove a diverse range of other compounds, including solvents, polymers, organic molecules, wastewater from kitchens, dairies, slaughterhouses, tanneries, petroleum industry, mining, biodiesel, edible oil refineries, tertiary pulp treatment, wood industries and municipal effluents, bacteria, oilseed, and olive-oil industries [29,30]. We noted an increase in the number of articles over time for all compounds evaluated (Table 1); however, this increase occurred to a greater extent for dyes (r = 0.85; p < 0.001), organic compounds (0.87, p < 0.001), pharmaceuticals (r = 0.81; p < 0.001), phenol (r = 0.81; p < 0.001), inorganic compounds/metals (r = 0.82, p < 0.001), pesticides (r =0.59; p < 0.001), hygiene and cleaning products (r = 0.42; p = 0.02), vinasse (r = 0.32; p = 0.09), and other types of specific composite products (r = 0.69; p < 0.001). These results are in agreement with the bibliometric analysis of WoS (see Figures S1 and S2 in the Supplementary Materials)

Regarding the treatment of wastewater with dyes, studies on this topic have increased in the last 10 years of the study period (2012–2022; r = 0.85 with time) [31,32,33]. Studies on the removal of organic compounds increased after 2015 (r = 0.87), showing that organic compounds were the most commonly used matrix in Fenton studies [34,35,36] in comparison with pharmaceuticals [37,38,39], phenols [40,41], pesticides [42,43], inorganic compounds [44,45], hygiene and cleaning products [46,47].

Figure 2 presents the sources of the pollutants listed in Table 1. However, 132 studies did not specify the origin of the treated matrix. Among those that were identified, 16 studies used the Fenton method to simultaneously degrade products from different sources, such as industrial effluent/laboratory-prepared solution (9), hospital effluent/laboratory-prepared solution (2), industrial effluent/other sources (1), industrial effluent/municipal wastewater treatment plant (WWTP) (1), and municipal WWTP/laboratory-prepared solution (2).

Most researchers have used Fenton for the degradation of industrial compounds (708; 56%), followed by laboratory-prepared solutions (synthetic effluent) (245; 19%), and other specific sources (98; 7.7%), such as sewage treatment plants (30; 2.4%), hospital sources (31; 2.5%), and research or educational institutions (3; 0.23%).

The volume of studies also correlates with the evolution and growth in the use of industrial processes and technologies over time, substantially increasing from the 2000s onward (Figure 1). Additionally, the search for new tools for sustainable industrial development involves introducing less environmentally harmful disposal techniques and then improving these techniques.

Most studies on Fenton reactions have been conducted on the laboratory-bench scale, representing 98.3% of the studies analyzed, followed by those on the pilot and industrial scales. On the laboratory-bench scale, a smaller volume of effluent is used under controlled laboratory conditions, even though these effluents are representative of industrial effluents, as shown in Figure 2.

Only 1.3% of the analyzed publications were conducted on the pilot scale, which is a method used for limiting the problems experienced with laboratory-bench-scale studies. However, pilot-scale studies are hindered by the operational costs associated with their implementation and execution. For example, Ferreira et al., (2020) [48] discussed the effect in a pilot plant process combining Fenton reaction with other AOPs and, then, the results showed that these kind of combinations are promising options for reducing the water footprint of biorecalcitrant industrial wastewaters. Moreover, 0.07% of direct studies have been conducted on the industrial scale, 0.07% of studies have used a combination of scales, and 0.31% of the studies did not specify the scale used.

Most studies (80.9%) used traditional reaction conditions (1022), followed by Fenton-like reactions (140) (Figure 3).

In the first case in Figure 3 on the y-axis, under ideal and optimized conditions, the compounds to be treated are catalytically oxidized in the presence of iron salts and hydrogen peroxide, in a pH range of 3.0 to 4.0. At values higher than 4.0, the rate of decomposition of hydrogen peroxide into water and oxygen markedly increases and the concentration of free iron species decreases due to the formation of iron complexes and precipitation of iron hydroxides. In the case of Fenton-like reactions, the catalyst can be other metals or complexes different from ferrous ion (Fe²⁺). Various monovalent and divalent cations and complexes have been considered, as noted by Jonshon and Mehrvar (2022). We found that 0.87% of articles reported the use of Fe²⁺ ions from secondary sources, such as industrial WWTP and mining sludge, whereas 1.3% of articles reported the use of Fe nanoparticles. The type of iron (hydro)oxide, whether natural or synthetic, solid, ionic, complex, or nano, affects the mechanism of ^•OH formation. We identified this information in this review; according to Figure 3, the use of Fenton-like reactions with cation exchange, nanoparticles, and the combination of these conditions in the same treatment has been increasing. Pereira et al. (2012) [49] discussed catalytic forms of iron oxide [50], as well as other metal nanoparticles [51] and complex ions [52].

For pH values below 3.0, the formation of hydroxyl radicals decreases, which becomes nonexistent at pH 1.0 [53]. Under these standard conditions, the reaction time is shorter and, consequently, the removal of compounds is lower compared with those of other AOPs [54]. Notably, only under acidic conditions does the ^•OH radical dominate, which favors the chemical oxidation process. Moreover, under normal reaction conditions, a certain amount of flocculated material accumulates, which is subsequently removed by decantation, due to the formation and complexation of iron salts [55] and the formation of highly valent oxidizing species, such as FeO³⁺ (Fe (V)) and the ferryl ion, FeO²⁺ (Fe (IV)) [56]. The concentration of hydrogen peroxide in the reaction must also be evaluated, as the excessive use of hydrogen peroxide in the reaction medium acts as a scavenger of hydroxyl radicals. The concentration of iron ions is also important for the process, ranging from 3 to 15 mg/L as, under the above conditions, the scavenging of hydroxyl radicals occurs, which decreases the efficiency of the reaction [57].

In a single study, different reactive means can be combined with the aim of comparing their efficacy. In this survey, we found that in 3.5% of the cases, the efficiencies of the traditional Fenton reaction, Fenton-like reactions, and parallel-shaped nanoparticles were tested in comparison with that of a single process.

Regarding the amount of organic matter removed, in the majority of studies, 884, more than 50% of the COD and TOC, was removed. However, only 30 studies reported that the Fenton treatment removed less than 50% of the COD and/or TOC, which was also reported in the literature [58]. For 349 studies, the reduction in COD and/or TOC was not reported.