A Systematic Review on Persulfate Activation Induced by Functionalized Mesoporous Silica Catalysts for Water Purification

Abstract

The eco-toxicological impacts caused by organic pollutants in aquatic environments have emerged as a global concern in recent decades, resulting from the potential hazards they present to ecosystem integrity and human health. Decorating active components on mesoporous silica is considered a popular approach by which to obtain synergistic effects in persulfate activation for sustainable water decontamination. However, at present there has been no review focusing solely, specifically and comprehensively on this field. Therefore, this paper places an emphasis on the latest research progress on the synthesis and physicochemical properties of functionalized mesoporous silica materials as well as their catalytic performance. The preparation methods included co-condensation, impregnation, grinding–calcination, hydrothermal synthesis and chemical precipitation, and their synthesis parameters played a major role in the characterization of materials, thereby affecting pollutant elimination. Metal redox cycles, nonmetallic activation and confinement effects contributed to persulfate activation. Targeted pollutants were degraded via radical pathways, non-radical pathways, or a combination of the two. The effects and causes of operational conditions (catalyst and persulfate dosage, initial pollutant concentration, temperature, initial pH, co-existing anions, and natural organic matter) varied across the degradation systems, and they were categorized and summarized in detail. Furthermore, functionalized mesoporous silica presented excellent reusability, stability and applicability in practical application. Finally, current potential directions for further research and sustainable development in this field were also prospected. This critical analysis aims to fuel the evolution of functionalized mesoporous silica catalyst-driven persulfate system application in water treatment and to bridge prevailing knowledge gaps.

1. Introduction

With the expansion of global industry and the rapid concentration of urban populations, environmental pollution has become a global challenge, one for which water pollution stands out as particularly persistent and which requires urgent global solutions [1,2]. Pollutants in water environments are diverse [3,4,5], and organic pollution has intensified to become one of the most concerning categories [6,7]. Global antibiotic consumption has increased significantly, with daily usage in Africa rising from 9.8 to 14.3 defined daily doses per 1000 inhabitants between 2000 and 2018, representing an increase of 46% [8]. Furthermore, pollution monitoring indicates the intensity of sulfonamide antibiotics usage reaching as high as 444.1–3160.5 mg/day/1000 inhabitants, calculated using the direct wastewater discharge point in Stellenbosch, South Africa [9]. The antibiotics were detected at a level of up to 92.3% in tap water samples in China, with macrolides and sulfonamides exhibiting the highest detection rates [10]. Water pollution caused by industrial emissions is equally severe. The textile industries in developing countries consumed 100,000–350,000 L water per ton of product, mainly for pretreatment, dye fixation and unreacted dyes removal. However, dye molecules are too difficult to be eliminated [11]. The concentrations of endocrine disrupting compounds (EDCs) showed a north–high, south–low distribution pattern with maximum levels reaching 50.9 ng/L in South Florida, USA [12], while they increased from the upstream to the estuary in the Bouregreg River in Morocco, among which 4-tert-octylphenol reached 368 ng/L and triclosan reached 301 ng/L [13]. The EDC pollution in the Dongjiang River Basin of China presented a decreasing distribution pattern from downstream to upstream, with total concentrations ranging from 149 to 2525 ng/L [14]. Regarding pesticide contamination, multiple herbicide residues were detected in the Aniema River in Nigeria, with peak concentrations of 0.43 μg/L for ivermectin and 0.33 μg/L for glyphosate [15]. In groundwater samples collected from 18 dairy farms across the Pampas Plain area of Córdoba Province, Argentina, atrazine (ATR) residues were detected in 50% of samples, with concentrations spanning 0.07–1.40 μg/L ($\overline{\mathrm{x}}$ = 0.25 ± 0.37). Nearly half (44.4%) exceeded the EU regulatory limit (0.1 μg/L) [16].

Aquatic organic pollutants pose multi-layered threats to both human health and ecosystems. Antibiotics exert toxic effects on humans and aquatic organisms. They not only inhibit protein expression and alter cell membrane permeability to affect bacterial growth, but also induce antibiotic resistance genes that significantly reduce microbial functional efficiency in ecosystems [17]. Carcinogenic substances in dye wastewater cause eye, throat, and skin irritation through drinking water or direct contact, resulting in asthma, allergic contact dermatitis and carcinogenic effects, and induce long-term harm to aquatic organisms [18]. EDCs interfere with endocrine functions, metabolism and the sexual development of mammals even at ng/L concentration [19]. Exposure of organisms to pesticides leads to metabolic disorders, organ damage, oxidative stress, and neurological damage, and also play adverse effects on the survival and development of algae, plankton, and benthic organisms, even resulting in individual death [20]. Consequently, the advancements of water remediation and wastewater treatment technology retain paramount significance as crucial measures by which to alleviate water resource crises and ensure long-term ecological health and sustainable development [21].

Traditional wastewater treatments have shown obvious limitations on the removal of pollutants characterized by persistence, recalcitrance, and low biodegradability. To confront this challenge, novel strategies have emerged in recent years, including chemical precipitation [22,23], ion exchange [24], electrochemical treatment [25], adsorption [26,27], bacteria/algae degradation [28,29], advanced oxidation processes (AOPs) [30,31,32,33], and membrane technology [34,35]. Among these, AOPs have garnered widespread attention due to their efficient removal of organic pollutants via the generation of highly reactive oxygen species (ROSs), and even complete mineralization without secondary pollution risks [36].

AOPs are classified with regard to homogeneous and heterogeneous catalysis. In homogeneous catalysis, the catalyst is dissolved as metal ions and coexists with reactants in the liquid phase. In spite of efficient catalysis, it is prone to problems of side reactions, unrecoverable catalysts, and secondary pollution from metal sludge [37]. In contrast, heterogeneous catalysis, which employs solid catalysts and avoids the aforementioned defects through solid–liquid interface reactions, has recently emerged as a dominant direction in the field of AOPs in terms of sustainable development [38]. Persulfate oxidation has garnered significant attention due to the notable advantages of sulfate radicals (SO₄^•−) possessing a higher redox potential, longer half-life, and high activity across a broad pH range [39,40]. Persulfuric acid is derived from Caro’s acid (H₂SO₅) and peroxydisulfuric acid (H₂S₂O₈) [41]. However, due to their instability in aqueous solution, the most commonly used persulfate oxidants are salt forms, specifically peroxydisulfate (PDS) and peroxymonosulfate (PMS). PDS typically exists as highly soluble and stable sodium or potassium salts. In contrast, PMS does not form stable mono-salts, with the exception of the triple salt (2KHSO₅·KHSO₄·K₂SO₄) [42]. Structurally, PDS contains two groups of SO₃⁻ and exhibits a symmetric configuration, which makes it more difficult to activate compared with the asymmetric PMS.

Despite plenty of heterogeneous catalysts reported for organic pollutants removal, conventional transition metallic catalysts have poor activation performance on persulfate due to the strong metal agglomeration and leaching tendency, which limits the long-term and efficient application in complex water environments [43]. Nonmetallic materials also do not exhibit excellent persulfate activation performance due to their weak electron conductivity, strong hydrophobicity, and inferior dispersion in water [44,45]. Mesoporous silica materials feature high specific surface area, large pore volume, long-range ordered channels, regular and adjustable pore size, easy surface functionalization, strong chemical stability, excellent biocompatibility, and hydrophilicity, and they are considered ideal supports for both metallic and nonmetallic modifications [46]. Furthermore, the confinement effect of functionalized mesoporous silica materials further contribute to pollutant removal [47].

However, current reviews concerned with functionalized mesoporous silica focus mainly on advances in synthesis methods [48], the adsorption removal of heavy metals [49] and perfluorinated compounds [50], adsorption and catalytic remediation for antibiotics [51,52], environmental remediation [53], and catalytic performance involving specific types of materials like Santa Barbar Amorphous (SBA) or metal-modified mesoporous silica [54,55,56]. To the best of our knowledge, there is no comprehensive review specifically addressing the activation of persulfate by various functionalized mesoporous silica materials for pollutant degradation from water. Therefore, this paper constructed four detailed content modules for detailed classification and elaboration, as follows: (1) synthesis methods and material characteristics of functionalized mesoporous silica; (2) removal efficiency and mechanisms of pollutants via persulfate activation by mono-metallic, bimetallic, nonmetallic and metallic–nonmetallic functionalized mesoporous silica materials; (3) influence effects of operational conditions (e.g., material and persulfate dosage, initial pollutant concentration) and water quality factors (e.g., temperature, initial pH, co-existing anions, and natural organic matter) on pollutant removal; and (4) regeneration, reusability, and stability of functionalized mesoporous silica materials, as well as their application potential in real water matrices. Possible future prospects with regard to challenges related to pollutant removal from water by functionalized mesoporous silica materials in persulfate activation were also proposed. This work aimed to integrate fragmented research findings and provide a comprehensive understanding of the progress in this field, in turn forming the basis for the future sustainable development of functionalized mesoporous silica materials in water purification for complex water pollution.

2. Synthesis Methods of Mesoporous Silica-Based Catalysts

The synthesis methods and material characteristics of functionalized mesoporous silica catalysts for persulfate activation are illustrated in Table 1, primarily including co-condensation, impregnation, grinding–calcination, hydrothermal synthesis, and chemical precipitation methods.

2.1. Co-Condensation Method

The co-condensation method is a common approach for preparing functionalized mesoporous silica. This process typically refers to the addition of metal precursors during the synthesis process of mesoporous silica, while simultaneously loading active species onto the silica [57]. This approach not only simplifies the preparation process, but also achieves highly dispersed metal active species.

Metal salts are commonly applied to be used as precursors in the co-condensation synthesis. To investigate the influence of different metal salts on material properties, cobalt (Co)-functionalized mesoporous silica materials were prepared using CoCl₂, Co(CH₃COO)₂, and Co(NO₃)₂ as precursors, designated as Co/SBA-15-Cl, Co/SBA-15-Ac, and Co/SBA-15-N, respectively [58]. Co primarily existed as Co₃O₄ with minor CoO content in Co/SBA-15-Cl and Co/SBA-15-Ac, where the Co₃O₄ crystallite size in Co/SBA-15-Cl (14.0 nm) exceeded that in Co/SBA-15-Ac (6.7 nm). Interestingly, a mixture of CoO, Co₂O₃, and Co₃O₄ was exhibited in Co/SBA-15-N, with Co₃O₄ crystallites being too small to be accurately measured. As the Co₃O₄ crystallite size in Co/SBA-15-Cl surpassed the pore diameter of SBA-15 (9 nm), most Co₃O₄ particles were located on the surface of SBA-15. Yang et al. [59] observed that MnCl₂ participated in reactions between the template and tetraethoxysilane (TEOS), where the MnCl₂ content affected the spherical morphology. Because the resulting manganese oxide species (MnO_x) were confined within mesopores, excessive MnCl₂ would increase the pore wall thickness and potentially block channels. Moreover, chelating agents could also be selected as precursors. Liu et al. [60] utilized sodium copper chlorophyllin (SCC) to prepare copper (Cu)-doped mesoporous silica (Cu-SBA-15), where Cu species existed in the form of highly dispersed single-atom Cu and CuO nanoparticles, accompanied by nonmetallic C and N. Zhang et al. [61] employed iron phthalocyanine to synthesize well-dispersed Fe₂O₃ nanoparticle-modified mesoporous silica (Fe-C/N-SBA-15). The formation of Fe-N and Fe-C coordination bonds facilitated Fe₂O₃ nanoparticle dispersion, creating isolated Fe species. The elevating Fe content led to an increase of specific surface area and pore size, due to the open pore structures maintained by the homogeneous distribution of Fe species. Metal nanoparticles are other precursor options. Xu et al. [62] used Fe₃O₄ particles as core and introduced a carbon layer to the area between the mesoporous silica and Fe₃O₄. High-temperature carbonization contributed to the conversation of Fe₃O₄ particles to Fe nanoparticles and simultaneously created void space for Fe nanoparticles, thereby constructing a core–shell structured Fe@void@mSiO₂.

Incorporating bimetallic precursors enables the synthesis of dual-metallic modified mesoporous silica materials. The Fe-Co-doped Mobil Composition of Matter No. 41 (MCM-41) catalyst (FeCo-MCM-41) was prepared using Fe(NO₃)₂ and Co(NO₃)₂ as precursors, and the results showed that the low metal doping did not affect the mesoporous order. Furthermore, Fe species entered the mesoporous silica framework and replaced the Si⁴⁺ to form Fe-O-Si bonds, which helped to maintain the stability of the catalyst [63]. Mn and Co incorporated into MCM-41 (Mn-Co-MCM-41) were synthesized using Mn((NO₃)₂ and Co(NO₃)₂ in a similar way [64]. Zhou et al. [65] conducted a study in which Co and Fe species were introduced into MCM-41 (γ-Fe₂O₃@2Co-MCM-41) through one-pot co-condensation using Fe₃O₄ and CoCl₂ as precursors of Fe and Co, respectively (Figure 1a). A large number of hydroxy groups were produced during the TEOS hydrolysis process, which promoted the nucleation of Fe₃O₄ micelles and enabled the Fe oxide species to be well encapsulated in the silica matrix. Fe₃O₄ was transformed into γ-Fe₂O₃ during the calcination process, while Co species incorporated the framework of mesoporous silica to form Co-O-Si bonds. In addition, the doping of Co enabled the suppression of γ-Fe₂O₃ aggregation and the improvement of their dispersion.

2.2. Two-Step Synthesis Method

The two-step synthesis method in the field of functionalized mesoporous silica synthesis is the widely used strategy, with focusing on the construct of mesoporous silica firstly and then followed by the introduction of active components [66,67]. This way includes impregnation, grinding–calcination, hydrothermal synthesis, and chemical precipitation methods.

2.2.1. Impregnation Method

The impregnation method involves immersing mesoporous silica in a solution containing metal precursors, followed by stirring, drying, or calcination to anchor active species onto the surface of mesoporous silica. This method preserves the original pore structure, morphology, and material property of mesoporous silicas [68,69]. However, it struggles to control the distribution of active species, particularly with metal nanoparticles prone to aggregation.

Metal nitrates are the most commonly used precursors in the impregnation method. Hu et al. [70] found that Co(II) of Co(NO₃)₂ was adsorbed on the surface of mesoporous silica via impregnation (Co/SBA-15), and was then pyrolysized to yield a stable Co-O-Si bond to alleviate the Co leaching. Co species existed as cubic spinel-phase Co₃O₄ and Co₂SiO₄ on the silica support, and the majority loaded on the pores of mesoporous silica, resulting in a decline in the specific surface area and pore volume. In contrast, Shao et al. [71] demonstrated that adding Co(NO₃)₂ to the solution containing the templated-SBA-16 ceramic coatings, followed by calcination, simultaneously removed templates and immobilized Co₃O₄ nanoparticle, as shown in Figure 1b. The obtained material was capable of higher specific surface area, pore volume, and better Co₃O₄ dispersion when compared with its traditionally impregnated counterpart (using template-free SBA-16).

Simultaneously impregnating two different metal nitrates with mesoporous silica is a widely used method for preparing bimetallic modified mesoporous silica materials, such as CoFe₂O₄-modified mesoporous silica (CF/SBA-15) [72], MnCo₂O₄-decorated mesoporous silica (CoMn/SBA-15) [73], Cu_0.76Co_2.24O₄-functionalized mesoporous silica (Cu_0.76Co_2.24O₄/SBA-15) [74], LaCoO_3-anchored mesoporous silica (LaCoO₃/SBA-15) [75], Cu_0.2Ni_0.8O-loaded mesoporous silica (Cu_0.2Ni_0.8O/SBA-15) [76], and Sr₂FeO₄-immobilized mesoporous silica (Sr₂FeO₄/SBA-15) [77]. Huang et al. [78] prepared a Co-based bimetallic doped with SBA-15 (AgCo/SBA-15) via impregnation, where the argentum (Ag) and Co species existed as Ag crystalline metal nanoparticles and Co₃O₄, respectively, facilitating electron transfer and preventing Co leaching. Similarly, Hu et al. [79] proposed that magnesium oxide (MgO) enhanced Co₃O₄ dispersion in Mg-doped Co-SBA-15 (CoMg/SBA-15) without forming large crystals.

With the exception of metal nitrates, other metal salts are also applied as precursors. Liu et al. [80] investigated the effects of different Co precursors on the materials’ properties. It was found that Co₃O₄ localized in the channels of mesoporous silicas (Co₃O₄@SBA-15) when using (CH₃COO)₂Co·4H₂O and Co(NO₃)₂·6H₂O as precursors, while Co₃O₄ loaded on the outer surface of mesoporous silica when utilizing CoCl₂·6H₂O. Yang et al. [81] synthesized MnO_x-doped SBA-15 (MnO_x/SBA-15) using Mn(CH₃COOH)₂·4H₂O, where Mn species co-existed as MnO, Mn₃O₄, Mn₂O₃, and MnO₂. Copper acetylacetonate (Cu(acac)₂) was also chosen to act as a Cu precursor to prepare Cu-MCM-41, but large CuO crystals primarily deposited on the external surface of MCM-41 [82].

The impregnation method also facilitates the synthesis of metallic–nonmetallic modified mesoporous silica. Co and N co-doped carbon composites (CoCNx@SBA-15) were synthesized utilizing vitamin B₁₂ as a Co, C and N precursor through impregnation and carbonization under a N₂ atmosphere [83], as the smaller molecular size of vitamin B₁₂ in comparison with the pore size of SBA-15 confined it to the channels of SBA-15. Yang et al. [84] impregnated SBA-15 with FeCl₃ and imidazole (Im) in ethanol, followed by N₂ calcination, yielding Fe-Im-SBA-15 with Fe-N bonds. This material exhibited a uniform distribution of Fe, C, and N within mesopores without aggregation, where Fe existed in intermediate valence states between Fe⁰ and Fe₂O₃ and carbon manifested as amorphous carbon with minor graphitic domains. In order to enhance metal loading capacity, Schlichter et al. [85] firstly prepared amino-modified mesoporous silica (NH₂-MCM-41) by the post-grafting method, and then leveraged amino–metal coordination with Cu coordinated to obtain the Cu/MCM-41-NH₂ composite.

2.2.2. Grinding–Calcination Method

The grinding–calcination method is an emerging environmentally friendly approach for synthesizing functionalized mesoporous silica catalysts [86]. This strategy involves thoroughly grinding mesoporous silica with active components or their precursors, followed by pyrolysis to immobilize the active species on mesoporous silica. This method demonstrates significant advantages through its simplified process, which consists of only two core steps: grinding and calcination. In terms of cost, this approach not only substantially reduces preparation expenses but also accommodates insoluble active components, overcoming the limitations inherent in traditional synthesis methods [87]. Regarding environmental benefits, it entirely avoids the use of organic solvents, minimizing the emission of harmful substances. Additionally, the relatively low calcination temperature effectively reduces energy consumption [88]. For industrial production, the straightforward process steps facilitate equipment integration and automation, supporting large-scale manufacturing. The low-temperature operation also lowers equipment requirements and investment costs, further enhancing the feasibility of industrial application [89].

The grinding–calcination method is widely used for synthesizing metallic-functionalized mesoporous silica catalysts, where surfactant-templated mesoporous silica plays a crucial role in the characteristics of active components during the grinding process. The confined space between silica walls and templates possess abundant silanol groups, which facilitates controlled insertion and uniformed distribution of metal ions during grinding. Subsequent calcination in air not only removes templates, but also enables dispersion of metal oxides inside mesoporous channels. Grinding Mn(NO₃)₂·4H₂O with the surfactant-templated SBA-15, followed by pyrolysis, yielded small Mn₂O₃ nanoparticles (7.1–8.5 nm) confined within mesopores [90]. Similarly, Khiem et al. [91] showed that Co₃O₄ grown and condensed on MCM-41 (Co@MCM-41) prepared by the grinding–calcination method achieved pore-confined Co₃O₄ dispersion, while the templated-free counterparts led to Co₃O₄ agglomeration on the external surface (Figure 1c). Yin et al. [92] reported grinding Co(NO₃)₂·6H₂O with SBA-15-containing templates followed by calcination generated a single Co site in SBA-15 (QS-CoS-SBA-15). In this composite, Co₃O₄ exhibited monolayer dispersion within mesopores without particle aggregation, forming quasi-single Co sites through Co-O-Si bonding. Additionally, grinding CoFe₂O₄ nanoparticles with surfactant-templated MCM-41 dispersed and immobilized CoFe₂O₄ on MCM-41 by pyrolysis, mitigating CoFe₂O₄ agglomeration [93].

This method also serves as an effective method for synthesizing nonmetallic functionalized mesoporous silica materials. Supporting carbon-based materials on mesoporous silica mitigates catalytic instability caused by carbon oxidation. Dicyandiamide, MCM-41 and distilled water were mixed and ground, followed by drying overnight for dicyandiamide recrystallization. The mixture was pyrolyzed at 550 °C in air to produce the MCM-41 containing graphite carbon nitride material (g-C₃N₄/MCM-41). The doped g-C₃N₄ on MCMC-41 led to both a decrease in the mesoporous order of MCM-41 and the crystallinity of g-C₃N₄. The length of g-C₃N₄ on g-C₃N₄/MCM-41 was approximately several micrometers, much shorter than that of bulk g-C₃N₄. This not only increased the specific surface area of g-C₃N₄, but also generated more edge defects and elevated the content of N-(C)₃ and N-C-O functional groups [94].

The metallic–nonmetallic bi-functionalized mesoporous silica catalyst could also be prepared by grinding and pyrolysis of metal salts, nonmetal precursors, and template-free mesoporous silica. Remarkably, unlike conventional methods using the templated mesoporous silica, this approach employed template-free mesoporous silica, pyrolyzed under a N₂ atmosphere to facilitate the incorporation of nonmetal elements. Currently, amino acids are the most used nonmetal precursors. Based on the confined effect of mesoporous silica channels and the inherent hydrophilic groups, the strong coordination of Co(II)/N or Co(II)/O was facilitated during the grinding process in order to integrate histidine (His) with mesoporous silica and His. When the calcination temperature was below 300 °C, Co species passed through ordered channels and were confined within the mesopores via capillary forces. As the calcination temperature increased to 600–900 °C, Co species were transformed into Co and CoO, while forming N/C layers within the pores. Therefore, the synergistic interaction among Co, mesoporous silica, and His yielded the magnetic Co-N/C-ordered mesoporous silica catalyst (Co@NC-ZS). Furthermore, increasing calcination temperature enhanced Co crystallinity, carbon content and defect density, and reduced N content [95]. Based on this work, sodium alginate (SA) encapsulation was employed to fabricate the Co/N-doped mesoporous silica combined with sodium alginate (Co/N@ZS-SA), where the alginate matrix was conducive to catalyst recovery [96]. Wang et al. [97] demonstrated that capillary forces facilitated the penetration of Fe(NO₃)₃·9H₂O and methionine (Met) into the mesopores of SBA-15 in a molten state during the grinding process (Figure 1d). Met acted as an essential prerequisite for driving Fe(NO₃)₃·9H₂O infiltration into silica channel. Otherwise, Fe(NO₃)₃·9H₂O tended to hydrolyze and aggregated on the external surface of SBA-15. Subsequent carbonization of Met generated C-N-S layers that coated the inner mesopores, while Fe(NO₃)₃·9H₂O was reduced to γ-Fe₂O₃ and subsequently sulfided into FeS nanoclusters, with S-containing species released by Met under the reductive environment. It was noted that strong coupling between Fe phases and C-N-S occurred through confined and chemical interactions without causing significant pore blockage or encapsulation. Remarkably, calcination temperature contributed to a critical determinant for controlling crystalline phase composition and particle dimension. Elevating the calcination temperature from 400 °C to 600 °C enlarged the particle sizes of γ-Fe₂O₃ and enhanced FeS crystallinity. Severe γ-Fe₂O₃ aggregation occurred, accompanied by FeS₂ formation at 700 °C. When the calcination temperature was further increased to 800 °C, continued γ-Fe₂O₃ growth was promoted and α-Fe₂O₃ emergence was triggered, with complete disappearance of Fe sulfide. This was due to the continuous depletion of C, S, and N elements caused by the low carbon yield of Met. Additionally, raising the Met/Fe(NO₃)₃·9H₂O ratio from 0.05 to 1.0 maintained γ-Fe₂O₃ as the dominant phase, though its diffraction intensity gradually weakened. This phenomenon stemmed from the following reasons: (1) the higher Met content suppressed Fe(NO₃)₃·9H₂O migration, (2) the generated C-N-S species improved γ-Fe₂O₃ dispersion, and (3) an elevated Met/Fe ratio promoted FeS formation.

2.2.3. Hydrothermal Synthesis

Hydrothermal synthesis is a technique for material synthesis in a sealed environment with solutions under high temperature and pressure. This method controls dissolution–recrystallization processes, yielding products with complete crystallization and good dispersion [98,99]. It is particularly effective for preparing metal-doped mesoporous silica catalysts. Liang et al. [100] added mesoporous silica to a solution containing Cu(NO₃)₂ and urea, where hydrolyzed urea reacted with Cu(NO₃)₂ to form Cu₂(OH)₂CO₃, which transformed into single-phase CuO nanoparticles with high crystallinity through calcination, as displayed in Figure 1e. The high specific surface area and abundant surface hydroxy groups of the mesoporous silica support provided numerous anchoring sites for CuO dispersion. The preparation of Cu-Al functionalized silica (Cu-Al/MSS) followed a similar procedure, where mesoporous silica effectively dispersed both CuO and Al₂O₃ particles and introduced Al₂O₃ increased oxygen vacancies [101]. Beyond preparing metal-doped mesoporous silicas, the hydrothermal synthesis method could also be applied to synthesize nonmetallic functionalized mesoporous silicas. Zhang et al. [102] synthesized NH₂-MCM-41 via an initial post-grafting, and then used hydrothermal methods to load graphene oxide (GO) onto NH₂-MCM-41 to obtain nitrogen-doped graphene-modified MCM-41 (NG-NH₂-MCM-41). This composite alleviated the wrinkling and stacking of nitrogen-doped graphene (NG), increased the effective active region, and enhanced hydrophilicity for better dispersion in water.

Figure 1. (a) Schematic diagram of the synthesis of (a) γ-Fe₂O₃@2Co-MCM-41 by co-condensation method, (b) Co-doped SBA-16 coated on ceramic monolith (Co@SBA-16/ceramic) through the impregnation method, (c) Co@MCM-41 and (d) Fe@C-N-S@SBA-15 obtained via the grinding–calcination method, and (e) CuO/MSS obtained through hydrothermal synthesis. Reproduced with permission from [65], [71], [91], [97] and [100] for (a), (b), (c), (d) and (e), respectively.

Figure 1. (a) Schematic diagram of the synthesis of (a) γ-Fe₂O₃@2Co-MCM-41 by co-condensation method, (b) Co-doped SBA-16 coated on ceramic monolith (Co@SBA-16/ceramic) through the impregnation method, (c) Co@MCM-41 and (d) Fe@C-N-S@SBA-15 obtained via the grinding–calcination method, and (e) CuO/MSS obtained through hydrothermal synthesis. Reproduced with permission from [65], [71], [91], [97] and [100] for (a), (b), (c), (d) and (e), respectively.

2.2.4. Chemical Precipitation

The chemical precipitation method adds mesoporous silica to a mixed solution containing metal ions, followed by alkaline precipitation to deposit metal species onto the surface of mesoporous silica. The functionalized mesoporous silica materials are then obtained after drying [103]. This method has the advantages of straightforward operation and low cost [104]. Huang et al. [105] immersed SBA-15 into a solution containing FeCl₃ and FeSO₄, then added ammonia to form Fe₃O₄ nanoparticles deposited on SBA-15, yielding Fe₃O₄ encapsulated in SBA-15 (Fe₃O₄@SBA-15). The Fe₃O₄ deposition did not influence the ordered pore structure of SBA-15 and possessed superparamagnetism. Compared with pure Fe₃O₄ particles (67 m²/g), Fe₃O₄@SBA-15 possessed higher surface area (241 m²/g). Although SBA-15 enhanced Fe₃O₄ dispersion and reduced their crystallite size, Fe₃O₄ particles were predominantly deposited on the external surface of SBA-15.

Table 1. Synthesis methods of functionalized mesoporous silica.

Catalyst	Synthesis Method	Specific Preparation Procedure	Material Characterization	Ref.
Co/SBA-15-Cl	Co-condensation	CoCl2, tri-block copolymer, and poly(ethylene glycol)-block-poly(propylene glycol)-block-poly (P123) were dissolved in HCl solution and TEOS was added. Solution was transferred to autoclave, washed, dried, and calcined in air to remove template.	DCo3O4 = 14.0 nm, d = 9 nm, PS = 1–3 μm	[58]
MnOx-incorporated mesoporous silica (Mn/MSS)		Dodecylamine (DDA) was dissolved in a water and ethanol mixture, which was stirred and MnCl2·4H2O was added until color became brown translucent. TEOS was added and the solution was centrifuged and calcined in air.	S = 800 m2/g, Vtp = 0.54 cm3/g, d = 2.14 nm, Mn = 6.2 wt%, a0 = 4.70 nm, d(100) = 4.07 nm	[59]
Cu-SBA-15		P123 and SCC were dissolved and added to HCl solution to form micelles. TEOS was added, and the solution was stirred, transferred to a Teflon-lined autoclave and calcined in air.	S = 1198.9 m2/g, Vtp = 0.8 cm3/g, d = 1.1 nm, PS = 3.5 μm, total Cu = 0.68 wt%, d(002) = 0.22 nm, Cu = 0.22 wt%, N = 0.26 wt%, Si = 12.14 wt%, O = 22.56 wt%, C = 64.82 wt%	[60]
Fe-C/N-SBA-15		P123 and iron phthalocyanine were mixed in HCl solution to form micelles. TEOS was added and the solution was stirred, transferred to a Teflon-lined autoclave, and calcined in air.	Fe = 2.51 wt%, N = 0.31 wt%, Si = 15.19 wt%, O = 26.46 wt%, C = 55.52 wt%, Dα-Fe2O3 = 20–40 nm, S = 1132.2 m2/g, Vtp = 0.8 cm3/g, d = 3.4 nm, d(104) = 0.27 nm	[61]
Fe@void@mSiO2		An Fe3O4 ethanol solution was ultrasonically dispersed in a water and ethanol mixture. Formaldehyde, resorcinol and NH3·H2O were sequentially added to form Fe3O4@RF microspheres. Dispersed Fe3O4@RF nanoparticles in hexadecyltrimethylammonium bromide (CTAB) solution, added cyclohexane and TEOS, stirred, washed, and calcined under N2.	Fm = 105 emu/g, S = 495 m2/g, d = 6.9 nm	[62]
FeCo-MCM-41		TEOS and ethanol solution were added to the mixture of CTAB, NaOH and water to form gel, Fe(NO3)3 and Co(NO3)3 were introduced and the solution placed into an autoclave, filtered, dried and calcined.	d(100) = 4.34 nm, a0 = 5.01 nm, S = 805 m2/g, d = 3.64 nm, Vtp = 0.80 cm3/g, Fe = 1.05 wt%, Co = 1.33 wt%	[63]
Mn-Co-MCM-41		A thermosensitive polymer was dissolved in water; pH = 11 was adjusted with ammonia; ethyl orthosilicate, Mn(NO3)2 and Co(NO3)2 were added; and the solution was placed in a sealed high-pressure vessel, centrifuged and ground, and calcined under N2.	d = 2 nm, DH = 39.5 nm, PDI = 0.227	[64]
γ-Fe2O3@2Co-MCM-41		CTAB was dissolved in water, CoCl2·6H2O and pre-synthesized Fe3O4 micelles were added, pH was adjusted to 10 with NH3·H2O, then TEOS was added to form a gel. The product was washed, dried, and calcined in air.	S = 644 m2/g, Vtp = 0.66 cm3/g, Co = 3.10 wt%, Fe = 3.23 wt%, a0 = 4.82 nm, T = 2.05 nm, d = 2.77 nm, d(100) = 4.17 nm	[65]
Co/SBA-15	Impregnation	An aqueous solution of Co(NO3)2·6H2O and SBA-15 was impregnated, dried, calcined in air, and ground.	DCo3O4 = 15.3 nm, S = 587.0 m2/g, Vtp = 0.717 cm3/g, d = 6.2 nm	[70]
Co@SBA-16/ceramic		A P123-templated SBA-16/ceramic was impregnated with Co(NO3)2 solution, dried, and calcined.	d = 3.1 nm, S = 7.6 m2/g, Vtp = 0.007 cm3/g, Co = 0.9 wt%	[71]
CF/SBA-15		Co(NO3)2·6H2O and Fe(NO3)3·9H2O were dissolved in water, mixed with SBA-15, dried, and calcined in air.	S = 506.1 m2/g, Vtp = 0.669 cm3/g, d = 6.6 nm, Co = 9.33 wt%, Fe = 8.93 wt%, Fm = 8.3 emu/g, T = 4.95 nm, DCo-Fe composites = ~100 nm	[72]
CoMn/SBA-15		SBA-15 was impregnated with aqueous Co(NO3)2 and Mn (NO3)2, dried, and calcined.	S = 452 m2/g, Vtp = 0.77 cm3/g, d = 8.26 nm, T = 13.5 nm	[73]
Cu0.76Co2.24O4/SBA-15		SBA-15 was mixed with Cu(NO3)2·3H2O and Co(NO3)2·6H2O in suspension, dried, and calcined in air.	S = 320.1 m2/g, Vtp = 0.8456 cm3/g, d = 5.2765 nm, pHpzc = 7.8	[74]
LaCoO3/SBA-15		SBA-15 was added to water, while La(NO3)3·6H2O and Co(NO3)2·6H2O were dissolved in ethanol with citric acid. The solution was dripped into SBA-15, stirred to form a gel, and calcined in air.	S = 99.19 m2/g, d = 6.78 nm, Vtp = 0.17 cm3/g, d(104) = 0.26 nm	[75]
Cu0.2Ni0.8O/SBA-15		Ni(NO3)2·6H2O and Cu(NO3)2·3H2O were dissolved in water under ultrasound, mixed with SBA-15, evaporated to a paste, dried, and calcined in air.	S = 320.52 m2/g, Vtp = 0.8456 cm3/g, d = 10.553 nm, Cu = 0.75 at%, Ni = 1.31 at%, O = 66.35 at%, Si = 31.59 at%, d(111) = 0.242 nm	[76]
Sr2FeO4/SBA-15		Sr(NO3)2 and Fe(NO3)3·5H2O were dissolved in water, mixed with SBA-15 to form a colloidal precursor, calcined in air, and ground.	d(103) = 0.2842 nm	[77]
AgCo/SBA-15		SBA-15 was impregnated with aqueous Co(NO3)2·6H2O and AgNO3, dried, and calcined in air.	S = 774 m2/g, Vtp = 1.077 cm3/g, d = 6.505 nm, Co = 3.0 wt%, C = 0.43 wt%, O = 54.26 wt%, Si = 39.52 wt% Co3O4: d(400) = 0.202 nm, d(222) = 0.233 nm, d(311) = 0.244 nm Ag: d(111) = 0.217 nm, Co/Si = 4.88 wt%, Ag = 0.91 wt%	[78]
CoMg/SBA-15		Mg/SBA-15 was prepared by impregnating SBA-15 with Mg(NO3)2·6H2O, then Co(NO3)2·6H2O was impregnated onto Mg/SBA-15, dried, and calcined in air.	S = 334.5 m2/g, Vtp = 0.422 cm3/g, d = 5.0 nm, DCo and Mg oxide crystallites = 6.6 nm	[79]
Co3O4@SBA-15		SBA-15 was added to a three-necked flask, (CH3COO)2Co·4H2O and urea were dissolved in ethylene glycol, dripped into the flask, transferred to an autoclave, and calcined in ethylene glycol.	S = 241 m2/g, Vtp = 0.62 cm3/g, Co = 0.45 at%, Vm = 0.008 cm3/g, d = 8.18 nm	[80]
MnOx/SBA-15		Mn(CH3COOH)2·4H2O and SBA-15 were ultrasonically impregnated in water, then calcined in air.	d = 5.54–6.65 nm, pHpzc = 2.70, S = 416.5 m2/g, Vtp = 0.73 cm3/g, Mn(IV) = 31.84 at%, Mn(III) = 20.72 at%, Mn(II) = 47.44 at%	[81]
Cu-MCM-41		MCM-41 was mixed with Cu(acac)2 in toluene, stirred, dried, and calcined in air.	S = 864 m2/g, Cu = 18.0 wt%, a0 = 4.15 nm	[82]
CoCNx@SBA-15		SBA-15 and vitamin B12 aqueous solution were mixed with CCl4 and n-butanol (to remove residual VB12 on SBA-15 surface), filtered, dried, and carbonized under N2.	S = 585 m2/g, Vtp = 0.94 cm3/g, d = 6.1 nm, Co = 0.32 wt%, C = 5.11 wt%, H = 1.37 wt%, N = 0.88 wt%, Vm = 0.035 cm3/g	[83]
Fe-Im-SBA-15		SBA-15, FeCl3, and imidazole were dispersed in ethanol, stirred, evaporated, calcined under N2, washed, and dried.	S = 294.27059 m2/g, Vtp = 0.44997 cm3/g, d = 6.01384 nm, Fe = 0.23 wt%, N = 0.68 wt%	[84]
Cu/MCM-41-NH2		MCM-41 was refluxed with toluene and 3-aminopropyltriethoxysilane (APTES) under N2, evaporated, washed with CHCl3 to obtain MCM-41-NH2, then mixed with CuSO4 solution.	Cu = 14.0 wt%, S = 9 m2/g	[85]
Mn2O3-anchored SBA-15 (Mn/asSBA-15)	Grinding–calcination method	Mn(NO3)2·4H2O was ground with templated-SBA-15 and calcined in air.	DMn2O3 = 7.1 nm, S = 642 m2/g, Vtp = 0.845 cm3/g, Mn = 3.03 mol%, d = 20 nm	[90]
Co@MCM-41		Co(NO3)2·6H2O was ground with CATB-templated MCM-41, ground, and calcined in air.	S = 262 m2/g, Vtp = 0.145 cm3/g, d = 2 nm, d(220) = 0.291 nm, d(200) = 0.393 nm	[91]
QS-CoS-SBA-15		Co(NO3)2·6H2O was ground with surfactant-templated SBA-15 and then calcined in air.	a0 = 11.3 nm, S = 531 m2/g, Vtp = 0.766 cm3/g, Co = 3.01 mmol/g, d = 6.1 nm	[92]
CoFe2O4-modified MCM-41 (CoFe2O4-MCM-41)		Templated-MCM-41 was ground with MCM-41, and then calcined in air.	S = 395.82 m2/g, Vtp = 0.4151 cm3/g, Fm = 23.05 emu/g, Fe = 5.13 at%, Co = 2.86 at%, Si = 21.88 at%, O = 70.12 at%, d(111) = 0.48 nm, d(220) = 0.295 nm	[93]
g-C3N4/MCM-41		Dicyandiamide and MCM-41 were mixed in water, dried, crystallized, ground, and calcined in air.	S = 298 m2/g, Vtp = 0.329 cm3/g, d = 2.09 nm, g-C3N4 = 43.4 wt%	[94]
Co@NC-ZS		Co(NO3)2·6H2O, Zr-modified mesoporous silica, and histidine were mixed, ground, and calcined under N2 atmosphere.	S = 307.7 m2/g, Vtp = 0.36 cm3/g, d = 6.2 nm, N/C layer = 0.8 nm, C = 38.1 at%, N = 2.2 at%, O = 37.9 at%, Si = 20.7 at%, Zr = 0.6 at%, Co = 0.5 at%, ID/IG = 0.9476, graphitic N = 34.2%	[95]
Co/N@ZS-SA		Zr modified mesoporous silica, l-histidine, and Co(NO3)2·6H2O were ground, calcined under N2, then mixed with sodium alginate (SA).	S = 76.01 m2/g, d = 10.38 nm, Vtp = 0.2 cm3/g, C = 27.42 wt%, N = 3.50 wt%, O = 23.31 wt%, Co = 9.02 wt%, Si = 25.63 wt%, Zr = 0.97 wt%, ID/IG = 1.32, C = 44.6 at%, N = 4.89 at%, O = 28.46 at%, Co = 2.99 at%, Zr = 0.21 at%, Si = 17.83 at%	[96]
FeS/γ-Fe2O3@N/S-doped SBA-15 (Fe@C-N-S@SBA-15)		SBA-15, Fe(NO3)3·9H2O, and Met were ground, and calcined under a N2 atmosphere.	Dγ-Fe2O3 = ~7.2 nm, d = 8.6 nm, S = 321.5 m2/g, Vtp = 0.64 cm3/g, C-N-S = 4.0 wt%, Fe = 8.3 wt%, C = 3.68 wt%, N = 0.41 wt%, S = 2.90 wt%	[97]
CuO-loaded mesoporous silica spheres (CuO/MSS)	Hydrothermal synthesis	Cu(NO3)2 and urea solutions were mixed with MSS, ultrasonicated, heated to form green precipitates, washed, and calcined.	S = 21.41 m2/g, Vtp = 0.036 cm3/g, d = 3.42 nm, Cu = 57.74 wt%, -OH = 1.64 mM/g, θ = 30.74°	[100]
Cu-Al/MSS		Mesoporous silica was ultrasonically dispersed, mixed with Cu(NO3)2, Al(NO3)3, and urea solutions to form precipitates and calcined.	S = 83.00 m2/g, d = 4.95 nm, Vtp = 0.20 cm3/g, DCu–Al particles = 90 nm, pHpzc = 9.6	[101]
NG/NH2-MCM-41		NH2-MCM-41 was dispersed in water, mixed with GO solution, NH3·H2O, and hydrazine hydrate, stirred, hydrothermally reacted, filtered, washed, and dried.	d = 2.33 nm, S = 193.93 m2/g, N = 4.83 at%, C = 49.45 at%, O = 31.61 at%, Si = 14.11 at%, d(002) = 0.82 nm, pHpzc = 6.7, θ = 36.3°	[102]
Fe3O4@SBA-15	Chemical precipitation	FeCl3 and FeSO4 aqueous solutions were mixed, heated, added to NH3·H2O water with SBA-15 powder, and then they were ultrasonicated, washed to neutral pH, and dried.	S = 241 m2/g, d = 17.61 nm, d(311) = 0.2526 nm, DFe3O4= ~3.7 nm	[105]

Note: D: Particle diameter; PS: Particle size; d: Average pore diameter; d (crystal plane): Crystal plane spacing; S: Specific surface area; V_tp: Total pore volume; V_m: Micropore volume; a₀: Unit cell parameter; T: Thickness; F_m: Maximum magnetic intensity; pH_pzc: Point of zero charge; -OH: Surface hydroxyl density; θ: Contact angle; PDI: Polydispersity index; D_H: Hydrodynamic particle size; I_D/I_G: Intensity ratio of D band and G band.

Table 1. Synthesis methods of functionalized mesoporous silica.

Catalyst	Synthesis Method	Specific Preparation Procedure	Material Characterization	Ref.
Co/SBA-15-Cl	Co-condensation	CoCl2, tri-block copolymer, and poly(ethylene glycol)-block-poly(propylene glycol)-block-poly (P123) were dissolved in HCl solution and TEOS was added. Solution was transferred to autoclave, washed, dried, and calcined in air to remove template.	DCo3O4 = 14.0 nm, d = 9 nm, PS = 1–3 μm	[58]
MnOx-incorporated mesoporous silica (Mn/MSS)		Dodecylamine (DDA) was dissolved in a water and ethanol mixture, which was stirred and MnCl2·4H2O was added until color became brown translucent. TEOS was added and the solution was centrifuged and calcined in air.	S = 800 m2/g, Vtp = 0.54 cm3/g, d = 2.14 nm, Mn = 6.2 wt%, a0 = 4.70 nm, d(100) = 4.07 nm	[59]
Cu-SBA-15		P123 and SCC were dissolved and added to HCl solution to form micelles. TEOS was added, and the solution was stirred, transferred to a Teflon-lined autoclave and calcined in air.	S = 1198.9 m2/g, Vtp = 0.8 cm3/g, d = 1.1 nm, PS = 3.5 μm, total Cu = 0.68 wt%, d(002) = 0.22 nm, Cu = 0.22 wt%, N = 0.26 wt%, Si = 12.14 wt%, O = 22.56 wt%, C = 64.82 wt%	[60]
Fe-C/N-SBA-15		P123 and iron phthalocyanine were mixed in HCl solution to form micelles. TEOS was added and the solution was stirred, transferred to a Teflon-lined autoclave, and calcined in air.	Fe = 2.51 wt%, N = 0.31 wt%, Si = 15.19 wt%, O = 26.46 wt%, C = 55.52 wt%, Dα-Fe2O3 = 20–40 nm, S = 1132.2 m2/g, Vtp = 0.8 cm3/g, d = 3.4 nm, d(104) = 0.27 nm	[61]
Fe@void@mSiO2		An Fe3O4 ethanol solution was ultrasonically dispersed in a water and ethanol mixture. Formaldehyde, resorcinol and NH3·H2O were sequentially added to form Fe3O4@RF microspheres. Dispersed Fe3O4@RF nanoparticles in hexadecyltrimethylammonium bromide (CTAB) solution, added cyclohexane and TEOS, stirred, washed, and calcined under N2.	Fm = 105 emu/g, S = 495 m2/g, d = 6.9 nm	[62]
FeCo-MCM-41		TEOS and ethanol solution were added to the mixture of CTAB, NaOH and water to form gel, Fe(NO3)3 and Co(NO3)3 were introduced and the solution placed into an autoclave, filtered, dried and calcined.	d(100) = 4.34 nm, a0 = 5.01 nm, S = 805 m2/g, d = 3.64 nm, Vtp = 0.80 cm3/g, Fe = 1.05 wt%, Co = 1.33 wt%	[63]
Mn-Co-MCM-41		A thermosensitive polymer was dissolved in water; pH = 11 was adjusted with ammonia; ethyl orthosilicate, Mn(NO3)2 and Co(NO3)2 were added; and the solution was placed in a sealed high-pressure vessel, centrifuged and ground, and calcined under N2.	d = 2 nm, DH = 39.5 nm, PDI = 0.227	[64]
γ-Fe2O3@2Co-MCM-41		CTAB was dissolved in water, CoCl2·6H2O and pre-synthesized Fe3O4 micelles were added, pH was adjusted to 10 with NH3·H2O, then TEOS was added to form a gel. The product was washed, dried, and calcined in air.	S = 644 m2/g, Vtp = 0.66 cm3/g, Co = 3.10 wt%, Fe = 3.23 wt%, a0 = 4.82 nm, T = 2.05 nm, d = 2.77 nm, d(100) = 4.17 nm	[65]
Co/SBA-15	Impregnation	An aqueous solution of Co(NO3)2·6H2O and SBA-15 was impregnated, dried, calcined in air, and ground.	DCo3O4 = 15.3 nm, S = 587.0 m2/g, Vtp = 0.717 cm3/g, d = 6.2 nm	[70]
Co@SBA-16/ceramic		A P123-templated SBA-16/ceramic was impregnated with Co(NO3)2 solution, dried, and calcined.	d = 3.1 nm, S = 7.6 m2/g, Vtp = 0.007 cm3/g, Co = 0.9 wt%	[71]
CF/SBA-15		Co(NO3)2·6H2O and Fe(NO3)3·9H2O were dissolved in water, mixed with SBA-15, dried, and calcined in air.	S = 506.1 m2/g, Vtp = 0.669 cm3/g, d = 6.6 nm, Co = 9.33 wt%, Fe = 8.93 wt%, Fm = 8.3 emu/g, T = 4.95 nm, DCo-Fe composites = ~100 nm	[72]
CoMn/SBA-15		SBA-15 was impregnated with aqueous Co(NO3)2 and Mn (NO3)2, dried, and calcined.	S = 452 m2/g, Vtp = 0.77 cm3/g, d = 8.26 nm, T = 13.5 nm	[73]
Cu0.76Co2.24O4/SBA-15		SBA-15 was mixed with Cu(NO3)2·3H2O and Co(NO3)2·6H2O in suspension, dried, and calcined in air.	S = 320.1 m2/g, Vtp = 0.8456 cm3/g, d = 5.2765 nm, pHpzc = 7.8	[74]
LaCoO3/SBA-15		SBA-15 was added to water, while La(NO3)3·6H2O and Co(NO3)2·6H2O were dissolved in ethanol with citric acid. The solution was dripped into SBA-15, stirred to form a gel, and calcined in air.	S = 99.19 m2/g, d = 6.78 nm, Vtp = 0.17 cm3/g, d(104) = 0.26 nm	[75]
Cu0.2Ni0.8O/SBA-15		Ni(NO3)2·6H2O and Cu(NO3)2·3H2O were dissolved in water under ultrasound, mixed with SBA-15, evaporated to a paste, dried, and calcined in air.	S = 320.52 m2/g, Vtp = 0.8456 cm3/g, d = 10.553 nm, Cu = 0.75 at%, Ni = 1.31 at%, O = 66.35 at%, Si = 31.59 at%, d(111) = 0.242 nm	[76]
Sr2FeO4/SBA-15		Sr(NO3)2 and Fe(NO3)3·5H2O were dissolved in water, mixed with SBA-15 to form a colloidal precursor, calcined in air, and ground.	d(103) = 0.2842 nm	[77]
AgCo/SBA-15		SBA-15 was impregnated with aqueous Co(NO3)2·6H2O and AgNO3, dried, and calcined in air.	S = 774 m2/g, Vtp = 1.077 cm3/g, d = 6.505 nm, Co = 3.0 wt%, C = 0.43 wt%, O = 54.26 wt%, Si = 39.52 wt% Co3O4: d(400) = 0.202 nm, d(222) = 0.233 nm, d(311) = 0.244 nm Ag: d(111) = 0.217 nm, Co/Si = 4.88 wt%, Ag = 0.91 wt%	[78]
CoMg/SBA-15		Mg/SBA-15 was prepared by impregnating SBA-15 with Mg(NO3)2·6H2O, then Co(NO3)2·6H2O was impregnated onto Mg/SBA-15, dried, and calcined in air.	S = 334.5 m2/g, Vtp = 0.422 cm3/g, d = 5.0 nm, DCo and Mg oxide crystallites = 6.6 nm	[79]
Co3O4@SBA-15		SBA-15 was added to a three-necked flask, (CH3COO)2Co·4H2O and urea were dissolved in ethylene glycol, dripped into the flask, transferred to an autoclave, and calcined in ethylene glycol.	S = 241 m2/g, Vtp = 0.62 cm3/g, Co = 0.45 at%, Vm = 0.008 cm3/g, d = 8.18 nm	[80]
MnOx/SBA-15		Mn(CH3COOH)2·4H2O and SBA-15 were ultrasonically impregnated in water, then calcined in air.	d = 5.54–6.65 nm, pHpzc = 2.70, S = 416.5 m2/g, Vtp = 0.73 cm3/g, Mn(IV) = 31.84 at%, Mn(III) = 20.72 at%, Mn(II) = 47.44 at%	[81]
Cu-MCM-41		MCM-41 was mixed with Cu(acac)2 in toluene, stirred, dried, and calcined in air.	S = 864 m2/g, Cu = 18.0 wt%, a0 = 4.15 nm	[82]
CoCNx@SBA-15		SBA-15 and vitamin B12 aqueous solution were mixed with CCl4 and n-butanol (to remove residual VB12 on SBA-15 surface), filtered, dried, and carbonized under N2.	S = 585 m2/g, Vtp = 0.94 cm3/g, d = 6.1 nm, Co = 0.32 wt%, C = 5.11 wt%, H = 1.37 wt%, N = 0.88 wt%, Vm = 0.035 cm3/g	[83]
Fe-Im-SBA-15		SBA-15, FeCl3, and imidazole were dispersed in ethanol, stirred, evaporated, calcined under N2, washed, and dried.	S = 294.27059 m2/g, Vtp = 0.44997 cm3/g, d = 6.01384 nm, Fe = 0.23 wt%, N = 0.68 wt%	[84]
Cu/MCM-41-NH2		MCM-41 was refluxed with toluene and 3-aminopropyltriethoxysilane (APTES) under N2, evaporated, washed with CHCl3 to obtain MCM-41-NH2, then mixed with CuSO4 solution.	Cu = 14.0 wt%, S = 9 m2/g	[85]
Mn2O3-anchored SBA-15 (Mn/asSBA-15)	Grinding–calcination method	Mn(NO3)2·4H2O was ground with templated-SBA-15 and calcined in air.	DMn2O3 = 7.1 nm, S = 642 m2/g, Vtp = 0.845 cm3/g, Mn = 3.03 mol%, d = 20 nm	[90]
Co@MCM-41		Co(NO3)2·6H2O was ground with CATB-templated MCM-41, ground, and calcined in air.	S = 262 m2/g, Vtp = 0.145 cm3/g, d = 2 nm, d(220) = 0.291 nm, d(200) = 0.393 nm	[91]
QS-CoS-SBA-15		Co(NO3)2·6H2O was ground with surfactant-templated SBA-15 and then calcined in air.	a0 = 11.3 nm, S = 531 m2/g, Vtp = 0.766 cm3/g, Co = 3.01 mmol/g, d = 6.1 nm	[92]
CoFe2O4-modified MCM-41 (CoFe2O4-MCM-41)		Templated-MCM-41 was ground with MCM-41, and then calcined in air.	S = 395.82 m2/g, Vtp = 0.4151 cm3/g, Fm = 23.05 emu/g, Fe = 5.13 at%, Co = 2.86 at%, Si = 21.88 at%, O = 70.12 at%, d(111) = 0.48 nm, d(220) = 0.295 nm	[93]
g-C3N4/MCM-41		Dicyandiamide and MCM-41 were mixed in water, dried, crystallized, ground, and calcined in air.	S = 298 m2/g, Vtp = 0.329 cm3/g, d = 2.09 nm, g-C3N4 = 43.4 wt%	[94]
Co@NC-ZS		Co(NO3)2·6H2O, Zr-modified mesoporous silica, and histidine were mixed, ground, and calcined under N2 atmosphere.	S = 307.7 m2/g, Vtp = 0.36 cm3/g, d = 6.2 nm, N/C layer = 0.8 nm, C = 38.1 at%, N = 2.2 at%, O = 37.9 at%, Si = 20.7 at%, Zr = 0.6 at%, Co = 0.5 at%, ID/IG = 0.9476, graphitic N = 34.2%	[95]
Co/N@ZS-SA		Zr modified mesoporous silica, l-histidine, and Co(NO3)2·6H2O were ground, calcined under N2, then mixed with sodium alginate (SA).	S = 76.01 m2/g, d = 10.38 nm, Vtp = 0.2 cm3/g, C = 27.42 wt%, N = 3.50 wt%, O = 23.31 wt%, Co = 9.02 wt%, Si = 25.63 wt%, Zr = 0.97 wt%, ID/IG = 1.32, C = 44.6 at%, N = 4.89 at%, O = 28.46 at%, Co = 2.99 at%, Zr = 0.21 at%, Si = 17.83 at%	[96]
FeS/γ-Fe2O3@N/S-doped SBA-15 (Fe@C-N-S@SBA-15)		SBA-15, Fe(NO3)3·9H2O, and Met were ground, and calcined under a N2 atmosphere.	Dγ-Fe2O3 = ~7.2 nm, d = 8.6 nm, S = 321.5 m2/g, Vtp = 0.64 cm3/g, C-N-S = 4.0 wt%, Fe = 8.3 wt%, C = 3.68 wt%, N = 0.41 wt%, S = 2.90 wt%	[97]
CuO-loaded mesoporous silica spheres (CuO/MSS)	Hydrothermal synthesis	Cu(NO3)2 and urea solutions were mixed with MSS, ultrasonicated, heated to form green precipitates, washed, and calcined.	S = 21.41 m2/g, Vtp = 0.036 cm3/g, d = 3.42 nm, Cu = 57.74 wt%, -OH = 1.64 mM/g, θ = 30.74°	[100]
Cu-Al/MSS		Mesoporous silica was ultrasonically dispersed, mixed with Cu(NO3)2, Al(NO3)3, and urea solutions to form precipitates and calcined.	S = 83.00 m2/g, d = 4.95 nm, Vtp = 0.20 cm3/g, DCu–Al particles = 90 nm, pHpzc = 9.6	[101]
NG/NH2-MCM-41		NH2-MCM-41 was dispersed in water, mixed with GO solution, NH3·H2O, and hydrazine hydrate, stirred, hydrothermally reacted, filtered, washed, and dried.	d = 2.33 nm, S = 193.93 m2/g, N = 4.83 at%, C = 49.45 at%, O = 31.61 at%, Si = 14.11 at%, d(002) = 0.82 nm, pHpzc = 6.7, θ = 36.3°	[102]
Fe3O4@SBA-15	Chemical precipitation	FeCl3 and FeSO4 aqueous solutions were mixed, heated, added to NH3·H2O water with SBA-15 powder, and then they were ultrasonicated, washed to neutral pH, and dried.	S = 241 m2/g, d = 17.61 nm, d(311) = 0.2526 nm, DFe3O4= ~3.7 nm	[105]

Note: D: Particle diameter; PS: Particle size; d: Average pore diameter; d (crystal plane): Crystal plane spacing; S: Specific surface area; V_tp: Total pore volume; V_m: Micropore volume; a₀: Unit cell parameter; T: Thickness; F_m: Maximum magnetic intensity; pH_pzc: Point of zero charge; -OH: Surface hydroxyl density; θ: Contact angle; PDI: Polydispersity index; D_H: Hydrodynamic particle size; I_D/I_G: Intensity ratio of D band and G band.

3. Strategies for Achieving Homogeneous Distribution of Active Sites in Functionalized Mesoporous Silicas

The homogeneous distribution of active sites on functionalized mesoporous silica is a critical determinant of catalytic efficiency and stability in persulfate activation. Based on a comprehensive analysis of studies, the strategies for achieving high homogeneous active site distribution could be categorized into the following key aspects, as summarized from comparative studies in Table S1.

The presence of an ordered mesoporous carrier is paramount for preventing agglomeration. Direct comparisons between catalysts supported on mesoporous silica and their non-porous or carrier-free analogues unequivocally demonstrated the confinement effect. For example, Fe-Im-SBA [84], LaCoO₃/SBA-15 [75], Fe@C-N-S@SBA-15 [97], and Fe₃O₄@SBA-15 [105] showed homogeneous distributions of metal species, whereas their counterparts Fe-Im, LaCoO₃, Fe@C-N-S, and Fe₃O₄ were plagued by severe metal agglomeration due to the lack of mesopore limitation. Similarly, CuO/MSS [100] and CoFe₂O₄-MCM-41 [93] exhibited a homogeneous distribution of metal oxides compared with CuO/SS and CoFe₂O₄-SiO₂. The rigid pore walls of the mesoporous support physically confine the nucleation and growth of metal species, yielding highly dispersed nanoparticles.

The choice of synthesis method is another important factor. Table S1 indicates that the co-condensation method generally yielded superior homogeneous distribution of active sites compared with the conventional impregnation method. A homogeneous distribution of Fe and Co species was achieved in FeCo-MCM-41 synthesized via co-condensation. In contrast, its impregnation-prepared counterpart (FeCo-MCM-41-imp) suffered from channel clogging by metal oxides [63]. This highlights the advantage of incorporating metal species directly into the mesoporous silica matrix during the framework formation. Similar phenomena were observed in the γ-Fe₂O₃@2Co-MCM-41. Impregnation led to a deposition of metal precursors primarily at the pore openings or on the external surface, resulting in uneven distribution and pore blockage [65]. In contrast, the co-condensation method facilitated the incorporation of metal species directly into the framework formation mesoporous silica, promoting dispersion of metal species. Notably, special attention should be paid to the loading amount, or it will lead to the damage of mesoporous structure. Furthermore, the further development of more suitable synthesis methods is necessary.

A predominant factor governing homogeneousness is the optimization of the active site loading amount to match the confinement capacity of the mesoporous silica. Evidence consistently demonstrates that exceeding an optimal loading leads to the metal agglomeration or damage of the mesoporous structure. For instance, while Cu-MCM-41 (10 wt%) [82] and Co/SBA-15 (10 wt%) [70] exhibited homogeneous distributions of metal oxide crystals, their higher-loading counterparts Cu-MCM-41 (24 wt%) and Co/SBA-15 (12 wt%) suffer from the problem of metal oxide aggregation. This trend was also confirmed in CF/SBA-15 [72], MnO_x/SBA-15 [81], CFMCM-41 [106] and g-C₃N₄/MCM-41 [94]. When functionalized mesoporous silica was synthesized via the co-condensation method, an excessive loading of active species could lead not only to the aggregation of active sites but also to the disruption of the mesostructure, as observed in the case of Mn/MSS [59]. Therefore, calibrating the metal loading to match the specific surface area and pore volume of the support is a primary strategy by which to ensure high dispersion.

The use of as-synthesized mesoporous silica (containing retained templates) during the synthesis process is a highly effective strategy for achieving the homogeneous distribution of metal species. Catalysts synthesized with as-synthesized mesoporous silica, such as Co@SBA-16/ceramic [71], Mn/asSBA-15 [90], QS-CoS-SBA-15 [92], and Co@MCM-41 [91], exhibited markedly better dispersion compared with their template-free counterparts. The templates not only anchored metal precursors within the porous network, but also prevented their migration and agglomeration during the calcination process.

4. Application of Mesoporous Silica-Based Catalysts in Water Remediation

This section reviews the pollutant removal efficiency and mechanisms of modified mesoporous silica (including mono-metallic, bimetallic, nonmetallic, and metallic–nonmetallic doped mesoporous silica) through reactive species pathways during persulfate activation, as summarized in Table 2.

4.1. Metallic Functionalized Mesoporous Silica Materials

4.1.1. Monometallic Functionalized Mesoporous Silica Materials

Transition metals have garnered significant attention due to their excellent performance in activating persulfates [107]. However, challenges such as particle agglomeration and high metal ions leaching limit their practical applications. Transition metals anchored onto mesoporous silica not only mitigate agglomeration of metals species, but also relieve the leaching problem [93,108]. Among these, the field of Co-based functionalized mesoporous silica materials has become a research hotspot owing to those materials’ superior persulfate activation performance, and studies have reported that Co commonly existed as Co₃O₄ nanoparticles on their surface. Schlichter et al. [82] have demonstrated that Co₃O₄ supported on MCM-41 (Co-MCM-41) efficiently activated PMS, achieving a 99% orange G decolorization rate within 120 min and a 48% mineralization in 4 h. Shukla et al. [58] have suggested that the Co₃O₄ anchored on the external surface of SBA-15 exposed to PMS degraded phenol completely in 200 min via the generated SO₄^•− radical. In contrast, Liu et al. [80] have proposed that Co₃O₄ confined within SBA-15 pores drive PMS through a non-radical electron transfer pathway, realizing a 100% sulfamethoxazole (SMX) removal rate in 50 min. This efficiency outperformed the externally loaded Co₃O₄/SBA-15 (77.6%), Co₃O₄/SiO₂ (59.3%) and Co₃O₄ (60%) due to the short mass transfer distances and concentrated active species caused by the confinement effect [109]. Similarly, Khiem et al. [91] clarified that confining Co₃O₄ in MCM-41 pores enhanced dispersion and increased active site exposure, which favored PMS activation and in turn realized a nearly 100% azorubin S (AZRS) removal rate in 30 min by SO₄^•−, •OH, and ¹O₂, surpassing externally aggregated Co/MCM-41 and commercial Co₃O₄. Notably, immobilizing Co₃O₄-functionalized mesoporous silica onto 3D-printed ceramics with honeycomb channels improved fluid dynamics and mass transfer, and thus activated PMS to remove 78% of levofloxacin (LVF) in 180 min. A long-time continuous-flow system showed the removal efficiency of LVF decreased from approximately 90% to 55% within the first 240 min, thereafter stabilizing around 55%. ¹O₂ was considered as a dominated reactive species, and O₂^•−, SO₄^•− and •OH were also advantageous to LVF degradation. All of these were yielded in the PMS activation process, and the activation mechanisms are shown in Equations (1)–(8) [71]. Moreover, Yin et al. [92] have revealed that the formed Co-O-Si bonds realized atomic-level dispersion of Co, which was conductive to PMS activation due to the increased number of exposed active sites. Additionally, SO₄^•− and •OH, generated through the reactions of Co(II) with PMS, were found to be responsible for phenol removal, with complete degradation in 10 min (Equations (1)–(3) and (9)). The single-molecular Co-O-Si species involved in Co₃O₄, Co₂SiO₄ and CoO in Co/SBA-15 attained a more than 98% phenol removal rate in 120 min by the SO₄^•− produced in the PMS activation by Co(II). Furthermore, immobilizing Co/SBA-15 on hydrophilic polytetrafluoroethylene (PTFE) membranes enabled the prevention of the loss of powder catalysts, thereby facilitating recycling [70].

$${\mathrm{C}\mathrm{o}}^{2+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(1)

$${\mathrm{C}\mathrm{o}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{2+}+{\mathrm{H}}^{+}+\mathrm{S}{\mathrm{O}}_5^{•-}$$

(2)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{O}\to •\mathrm{O}\mathrm{H}+{\mathrm{H}}^{+}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(3)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}_2{\mathrm{O}}_2+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(4)

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

(5)

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

(6)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+\mathrm{S}{\mathrm{O}}_5^{2-}\to \mathrm{H}\mathrm{S}{\mathrm{O}}_4^{-1}+{}^1\mathrm{O}_2+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(7)

$${\mathrm{C}\mathrm{o}}^{2+}+\mathrm{S}{\mathrm{O}}_5^{•-}\to {\mathrm{C}\mathrm{o}}^{3+}+{}^1\mathrm{O}_2+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(8)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+• \mathrm{O}\mathrm{H}+{\mathrm{C}}_6{\mathrm{H}}_5\mathrm{O}\mathrm{H}\to \mathrm{s}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l} \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{p}\mathrm{s}\to \mathrm{C}{\mathrm{O}}_2+\mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{H}}_2\mathrm{O}$$

(9)

Cu-based functionalized mesoporous silica could drive persulfates via the Cu(II)/Cu(I) redox cycle, contributing to high-performance pollutant degradation [110,111]. The Cu-MCM-41 decolored 80% of orange G within 120 min, with 32% of decarburization after 6 h [82]. Sajjadi et al. [112] proposed that the highly dispersed CuO nanoparticles on Cu-doped mesoporous silica-based particles (Cu-BMS) were responsible for 93.5% of methylene blue (MB) removal in 60 min by SO₄^•− and •OH. Although both SO₄^•− and •OH were involved in the MB degradation, the quenching experiment indicated that SO₄^•− was the primary radical specie responsible for MB degradation in the Cu-BMS/PDS system. Liang et al. [100] suggested that Cu(II) in CuO/MSS formed hydrogen bonds with PDS to realize electron transfer, which was conductive to bisphenol A (BPA) degradation (92% in 45 min). The radical pathway only had a minor contribution to BPA removal. Liu et al. [60] demonstrated that the single-atom Cu sites formed tetrahedral structures with SBA-15 maximizing atomic utilization and providing high-density active sites, while CuO nanoparticles enabled Cu(II)/Cu(I) electron transfer. This synergistic system encouraged PMS to generate ¹O₂, •OH, SO₄^•−, and O₂^•− to remove 90% of tetracycline (TC) in 60 min, and the possible catalytic mechanism is shown in Equations (10)–(17).

$$\mathrm{C}{\mathrm{u}}^{2+}-\mathrm{O}\mathrm{H}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{C}{\mathrm{u}}^{2+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2\mathrm{O}$$

(10)

$$\mathrm{C}{\mathrm{u}}^{2+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}{\mathrm{u}}^{+}-\mathrm{O}\mathrm{H}+{\mathrm{H}}^{+}+\mathrm{S}{\mathrm{O}}_5^{•-}$$

(11)

$$\mathrm{C}{\mathrm{u}}^{+}-\mathrm{O}\mathrm{H}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{C}{\mathrm{u}}^{+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2\mathrm{O}$$

(12)

$$\mathrm{C}{\mathrm{u}}^{+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}{\mathrm{u}}^{2+}-\mathrm{O}\mathrm{H}+• \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{•-}$$

(13)

$$\mathrm{C}{\mathrm{u}}^{+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}{\mathrm{u}}^{2+}-\mathrm{O}\mathrm{H}+• \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(14)

$$\mathrm{C}{\mathrm{u}}^{2+}-\left(\mathrm{O}\right)\mathrm{O}\mathrm{S}{\mathrm{O}}_3^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}{\mathrm{u}}^{+}-\mathrm{O}\mathrm{H}+{2\mathrm{H}}^{+}+2\mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{O}}_2^{•-}$$

(15)

$${2\mathrm{O}}_2^{\mathrm{g}-}+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{H}}_2{\mathrm{O}}_2+{}^1\mathrm{O}_2+2\mathrm{O}{\mathrm{H}}^{-}$$

(16)

$$\mathrm{t}\mathrm{e}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{e}+{\mathrm{O}}_2^{\mathrm{g}-}/\mathrm{S}{\mathrm{O}}_4^{\mathrm{g}-}/•\mathrm{O}\mathrm{H}/{}^1\mathrm{O}_2\to \mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{s}\to \mathrm{C}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}$$

(17)

Mn possessing rich valence transitions exhibits superior redox activity in PMS application [113,114], and generally exists as mixed-valence MnO_x oxides on mesoporous silica. The MnO_x/SBA-15/PMS system eliminated 98.4% of butyl paraben (BPB) in 180 min via SO₄^•− and •OH, outperforming the MnO_x/SBA-15/PMS system due to the better dispersion [81]. Moreover, calcination temperature and duration critically influenced MnO_x valence states, thereby controlling radical speciation and concentration in MnO_x/SBA-15/PMS systems. Optimal material was obtained after calcination at 823.8 K for 3 h, where Mn₂O₃ and Mn₃O₄ (average Mn valence: +2.844) predominated, and these species exhibited superior PMS activation efficiency when compared with MnO₂. Yang et al. [59] confirmed MnO_x as a key active site in Mn/MSS for methyl orange (MO) degradation, and the MO removal efficiency presented an initial increase followed by a decline that was concomitant with increased Mn loading. This was mainly because the excessive Mn loading caused the agglomeration, change of Mn category, and blockage of mesoporous silica channels. The best-performing Mn/MSS featured diffusive mesoporous channels and the highest proportion of accessible MnO_x species (mainly Mn₃O₄), which could activate PMS effectively to produce SO₄^•− and •OH through the transformation process from Mn(II) and Mn(III) to Mn(IV) (Equations (18)–(25)), achieving a 90% MO removal rate within 8 min. Conversely, Lu et al. [90] believed the position and size of Mn₂O₃ on mesoporous silica affected the catalytic efficiency. When Mn loading was relatively low (1.0–2.0 mmol/g), the formed larger-sized Mn₂O₃ loaded on the outer surface of SBA-15 was more likely to activate PDS and degrade sulfachlorpyridazine (SCP) than the one with fine-sized Mn₂O₃ in the channel of SBA-15 under the same Mn loading conditions. As Mn loading was relatively high (3.0–5.0 mmol/g), the produced fine-sized Mn₂O₃ located in the pores of SBA-15 played a major role in PDS activation to obtain better SCP removal. Both the generated SO₄^•− and •OH participated in the SCP decomposition (Equations (3), (26) and (27)).

$$\mathrm{M}{\mathrm{n}}^{2+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{M}{\mathrm{n}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(18)

$$\mathrm{M}{\mathrm{n}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{M}{\mathrm{n}}^{4+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(19)

$$\mathrm{M}{\mathrm{n}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{M}{\mathrm{n}}^{4+}+\mathrm{S}{\mathrm{O}}_4^{2-}+• \mathrm{O}\mathrm{H}$$

(20)

$$\mathrm{M}{\mathrm{n}}^{4+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{M}{\mathrm{n}}^{3+}+\mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(21)

$$\mathrm{M}{\mathrm{n}}^{\left(3+/4+\right)}-\mathrm{O}\mathrm{H}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{M}{\mathrm{n}}^{\left(3+/4+\right)}-\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+\mathrm{O}{\mathrm{H}}^{-}$$

(22)

$$\mathrm{M}{\mathrm{n}}^{\left(3+/4+\right)}-\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+2{\mathrm{H}}_2\mathrm{O}\to \mathrm{M}{\mathrm{n}}^{\left(3+/4+\right)}-\mathrm{O}{\mathrm{H}}^{-}+\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{O}}_2+{4\mathrm{H}}^{+}$$

(23)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+• \mathrm{O}\mathrm{H}$$

(24)

$$2•\mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}}_2$$

(25)

$$\mathrm{M}{\mathrm{n}}^{2+}+{\mathrm{S}}_2{\mathrm{O}}_8^{2-}\to \mathrm{M}{\mathrm{n}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(26)

$$• \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{C}}_{10}{\mathrm{H}}_9\mathrm{C}\mathrm{l}{\mathrm{N}}_4{\mathrm{O}}_2\mathrm{S}\to \mathrm{s}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l} \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{p}\mathrm{s}\to \mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{C}\mathrm{O}}_2$$

(27)

Fe₃O₄ is also applied in functionalized mesoporous silica-activated persulfates, as a result of its low-cost advantage and magnetic property for convenient recovery. Huang et al. [105] discovered that the uniquely ordered mesoporous structure of SBA-15 was not only advantageous to carbamazepine (CBZ) and PDS adsorption in order to concentrate reactants at the interface, but also to the activation of PDS to generate SO₄^•−, O₂^•−, and •OH by highly dispersing Fe₃O₄, thereby promoting surface reactions. This system was available to realize a 100% CBZ degradation in 30 min at pH = 3.0 with 0.50 g/L of catalysts and 300 mg/L of PDS, higher than the Fe₃O₄/PDS system (78%). Liang et al. [115] discovered that removal capability was strongly correlated with surface wettability, where Fe₃O₄-modified mesoporous silica (Fe-MSS) possessed strong hydrophilicity to promote MB adsorption capacity, thus exhibiting the high MB removal rate. Furthermore, the Fe-MSS/PDS system effectively produced SO₄^•− radicals, achieving a 90% MB removal within 60 min. This was higher than the Fe₃O₄/PDS system (60%) due to the mesoporous structure preventing nanoparticle aggregation and the effect of confinement. The mesoporous silica shell in the core–shell Fe@void@mSiO₂ material had a relatively high adsorption rate (~40%) for TC, which was accumulated in the voids of the core–shell [62]. Subsequently, the Fe-activated PMS on the surface produced SO₄^•− and •OH to degrade the high concentration of TC in the confined space, attaining TC removal rate of 100% within 40 min.

4.1.2. Bimetallic Functionalized Mesoporous Silica Materials

Although mono-metallic modified mesoporous silica materials have demonstrated excellent catalytic performance in persulfate activation, bimetallic functionalized mesoporous silica catalysts have been found to further accelerate oxidation-reduction reactions through intermetallic synergistic effects [116], and to reduce amount of metals used while maintaining the comparable activity, thus reducing costs.

The Co-Fe bi-functionalized mesoporous silica is one of the most prevalent bimetallic systems for persulfate activation, typically CoFe₂O₄-modified mesoporous silica. The inherent magnetism of CoFe₂O₄ promotes facile magnetic separation from treated water and mitigates secondary ecological pollution risk from residual nanoparticles. Hu et al. [72] found that the Fe-Co bonds existing in CoFe₂O₄ nanoparticles on the inside and outer surfaces of mesoporous silica channels increased the number of hydroxy groups, which promoted the formation of Co(II)-OH complexes and then efficiently activated PMS to remove 98% of rhodamine B (RhB) in 120 min. Choong et al. [106] proposed that PMS was adsorbed onto CoFe₂O₄ doped in MCM-41 (CFMCM-41), and then activated by the Co(II)/Co(III) and Fe(II)/Fe(III) redox cycles to produce SO₄^•− and •OH to degrade more than 90% of ciprofloxacin (CIP) in 60 min. Gao et al. [93] indicated that the generated SO₄^•−, •OH, O₂^•−, ¹O₂ and electron transfer in the CoFe₂O₄-MCM-41-driven PMS system were responsible for the SMX degradation. Furthermore, a higher rate constant observed in the CoFe₂O₄-MCM-41/PMS system (0.10244 (min·mg·L⁻¹)⁻¹) than in the CoFe₂O₄-SiO₂/PMS system (0.00181 (min·mg·L⁻¹)⁻¹) suggested that the confinement effect could shorten mass transfer distances, concentrate reactive species, and facilitate efficient collisions with SMX, thereby accelerating mass/electron transfer and enhancing overall reaction kinetics. Beyond conventional CoFe₂O₄-loaded mesoporous silica, a core–shell catalyst (CFS) with inverse spinel CoFe₂O₄ as a core and mesoporous silica as a shell decomposed acid orange II with more than 90% removal capability in 35 min, outperforming both SiO₂-supported catalyst and naked CoFe₂O₄ nanoparticles [117]. Notably, some research has revealed that Fe and Co exist as non-spinel Co-Fe configurations in Fe-Co bimetallic modified mesoporous silicas. Sun et al. [63] proposed that, although the Co content in FeCo-MCM-41 was only half of that in Co-MCM-41, its efficacy in PMS activation and MO degradation was comparable to that of Co-MCM-41, highlighting the synergistic effect of the Fe-Co bimetallic system. The hybrid mechanism of PMS activation is summarized in Equations (28)–(39). Activated CoOH⁺ was initially formed, and then the reaction occurred between the adsorbed PMS and CoOH⁺/Fe(II) to produce SO₄^•−. ¹O₂ was yielded through multiple pathways, including the reactions of SO₅^•− with H₂O, PMS self-decomposition, and transformation of chemisorbed oxygen (O*) to the lattice oxygen. The regeneration of Fe(II) and Co(II) was achieved via the reduction of Fe(III)/Co(III) by PMS. In addition, Fe(II) was conducive to the generation of Co(II). Zhou et al. [65] demonstrated that γ-Fe₂O₃ embedded in an MCM-41 framework could produce a synergistic action with its adjacent Co to activate PMS and degrade 99% of acid orange II within 40 min. Its performance was superior to that of γ-Fe₂O₃ confined within MCM-41, Co₃O₄ supported on MCM-41, and the bi-functional material combining both. It is worth noting that the bi-functionalized material exhibited a similar catalytic performance to the Co₃O₄ decorated on the surface of MCM-41, indicating no synergistic effect between the surface Co₃O₄ and matrix-confined γ-Fe₂O₃ in MCM-41.

$$\mathrm{C}\mathrm{o}{\left(\mathrm{O}\mathrm{H}\right)}_2+{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{o}\mathrm{O}\mathrm{H}}^{+}+{\mathrm{H}}_2\mathrm{O}$$

(28)

$${\mathrm{C}\mathrm{o}\mathrm{O}\mathrm{H}}^{+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}\mathrm{O}}^{+}+\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{O}$$

(29)

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

(30)

$${\mathrm{F}\mathrm{e}}^{2+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{F}\mathrm{e}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(31)

$${\mathrm{C}\mathrm{o}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{2+}+\mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(32)

$${\mathrm{F}\mathrm{e}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{F}\mathrm{e}}^{2+}+\mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(33)

$$2\mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}_2\mathrm{O}\to 1.5{}^1\mathrm{O}_2+2\mathrm{H}\mathrm{S}{\mathrm{O}}_4^{-}$$

(34)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{S}{\mathrm{O}}_5^{2-}+{\mathrm{H}}^{+}$$

(35)

$$\mathrm{S}{\mathrm{O}}_5^{2-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to 2\mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{H}}^{+}+{}^1\mathrm{O}_2$$

(36)

$${\mathrm{O}}_{\mathrm{L}\mathrm{a}\mathrm{t}}\to {\mathrm{O}}^{\mathrm{*}}$$

(37)

$${\mathrm{O}}^{\mathrm{*}}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {}^1\mathrm{O}_2+\mathrm{H}\mathrm{S}{\mathrm{O}}_4^{-}$$

(38)

$${\mathrm{C}\mathrm{o}}^{3+}+{\mathrm{F}\mathrm{e}}^{2+}\to {\mathrm{C}\mathrm{o}}^{2+}+{\mathrm{F}\mathrm{e}}^{3+}$$

(39)

Co is considered as a dominant active species in Co-Mn bi-functionalized mesoporous silica catalysts, with Mn promoting Co(II) regeneration through multi-valent electron transfer. The RhB removal rates were less than 30% in both the Co/SBA-15/PMS and Mn/SBA-15/PMS systems, whereas the CoMn/SBA-15/PMS system removed RhB of 99% in 45 min. This indicated a significant synergistic catalytic effect between Mn and Co, where they loaded on SBA-15 in the form of MnCo₂O₄. The activation mechanism of CoMn/SBA-15 first involved H₂O dissociation to form Co-OH and Mn-OH, and then adsorbed PMS to generate ROSs, such as SO₄^•−, •OH, and ¹O₂ [73]. Peng et al. [64] also discovered that Mn-Co-MCM-41 encouraged PMS to produce SO₄^•−, •OH, and O₂^•− to remove 98% of RhB in 20 min.

Beyond Co-Fe and Co-Mn modified mesoporous silicas, Co could also form with diverse metals, such as Cu, Ag, La, and Mg. The incorporation of Cu(I) thermodynamically facilitated the Co(II)-Co(III)-Co(II) redox cycling during PMS activation (Equations (1), (2) and (40)–(42)), which presented a 99.2% sulfapyridine (SPD) removal rate in 90 min, higher than that in the CuO/SBA-15/PMS (<50%) and Co₃O₄/SBA-15/PMS (76%) systems [74]. The Ag in Ag-Co/SBA-15 acted as an electron bridge to favor the reduction of Co(III) to Co(II). Meanwhile, Ag possessed the antibacterial property that prevented biofouling and ensured long-term stability in complex water bodies. Comparative studies carried out on unsymmetrical dimethylhydrazine (UDMH) degradation by PMS revealed distinct catalytic efficiencies in the order of AgCo/SBA-15 (0.1602 min⁻¹) > CuCo/SBA-15 (0.1295 min⁻¹) > Co/SBA-15 (0.1177 min⁻¹) > CeCo/SBA-15 (0.0977 min⁻¹) > FeCo/SBA-15 (0.0912 min⁻¹), suggesting that the incorporation of Ag and Cu into Co/SBA-15 provided superior catalytic performance. Further analysis demonstrated that the introduced Ag increased UDMH adsorption despite minor porosity reduction, and that its high conductivity accelerated redox cycling between Co(II)/Co(III). This resulted in complete UDMH degradation through combined radical (SO₄^•− and •OH) and non-radical (¹O₂) pathways, and ¹O₂ played a major role [78]. Hu et al. [79] found that MgO in the CoMg/SBA-15 increased surface alkalinity and hydroxy group density and facilitated Co(II)-OH complex formation, which enhanced SO₄^•− generation. MgO also improved Co₃O₄ dispersion on SBA-15, preventing active site aggregation. This led to a 100% RhB removal rate within 5 min, dominantly achieved via SO₄^•− pathways. Similarly, LaCoO₃/SBA-15 prompted Co(II)-Co(III)-Co(II) redox cycling through La(III)/La(IV) electron transfer, activating PMS to produce ¹O₂, SO₄^•−, •OH, and O₂^•− for ATR degradation, with a near 100% degradation in 6 min [75]. This removal rate was higher than the LaCoO₃/PMS system (55.15%) due to the short, ordered pathways for reactant diffusion.

$${\mathrm{C}\mathrm{u}}^{+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{u}}^{2+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(40)

$${\mathrm{C}\mathrm{u}}^{+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{u}}^{2+}+\mathrm{S}{\mathrm{O}}_5^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(41)

$${\mathrm{C}\mathrm{o}}^{3+}+{\mathrm{C}\mathrm{u}}^{+}\to {\mathrm{C}\mathrm{o}}^{2+}+{\mathrm{C}\mathrm{u}}^{2+}$$

(42)

Apart from Co-based bimetallic-supported mesoporous silicas, the co-doping of other bi-metals on mesoporous silicas also demonstrated significantly enhanced catalytic effects. The redox pair active sites formed by the Cu-O-Ni bonds in pure cubic Cu_0.2Ni_0.8O nanoparticles decorated on SBA-15 promoted a Ni(III)/Ni(II) cycle through the Cu(I)/Cu(II) cycle. This strategy markedly boosted PMS activation to produce O₂^•−, SO₄^•−, •OH, and ¹O₂ for SCP removal (99.27% in 90 min) [76]. Wang et al. [101] identified that the introduction of Al₂O₃ in Cu-Al/MSS increased the oxygen vacancy content and facilitated persulfate–catalyst ligand complexation (Cu(II)-O-SO₃^•). The electron transfer between electron donor BPA and electron acceptor Cu(II)-O-SO₃^• governed BPA degradation to result in the removal of 90% in tin 30 min. In addition, the formed Sr-O-Fe bond in Sr₂FeO₄/SBA-15 was capable of a stable cycle of Fe(II)-Fe(III)-Fe(II) through the electron-rich Sr. This catalyst was available to activate PMS to remove SPD of 99% in 90 min by produced SO₄^•− and •OH [77].

4.2. Nonmetallic Functionalized Mesoporous Silica Materials

While pure carbon materials present slow electron transfer, doping with heteroatoms such as N or S could effectively enhance the electron transfer capability. However, the inherent hydrophobicity of carbon-based carriers often results in poor dispersion in aqueous solution. The introduction of hydrophilic mesoporous silica is an effective strategy to mitigate this limitation. Dong et al. [94] demonstrated that oxygen groups on g-C₃N₄/MCM-41 served as PMS activation sites. The electron-rich oxidized carbon species N-C-O was recognized as an intermediate, which was oxidized by PMS to N-C=O and then reduced back to C-OH to complete the catalytic cycle. This process generated SO₄^•−, ·OH, and non-radical oxidation to degrade acid orange 7 (AO7) with 96.8% removal in 30 min. The g-C₃N₄/MCM-41/PMS system also efficiently degraded diverse dyes (reactive brilliant red X-3B (X-3B), MO, RhB, MB, reactive brilliant blue KN-R (KN-R)) within 30 min. In addition, p-cresol (p-CR) and BPA were adsorbed on NG/NH₂-MCM-41 by π–π and electrostatic interactions, and PMS was driven by the N-doped regions of NG/NH₂-MCM-41 simultaneously to produce SO₄^•−, •OH, and ¹O₂, resulting in an approximately 100% BPA and 95% p-CR removal in 700 min. The SO₄^•− and •OH were generated by electron transfer activation of PMS (Equations (43) and (44)), and ¹O₂ was produced by self-decomposition of PMS (Equation (7)) and radical processes (Equations (34) and (45)). Because BPA had more π electrons than p-CR, it was more easily adsorbed onto the surface of NG/NH₂-MCM-41. Moreover, the increase in the number of phenolic hydroxy groups and benzene rings led to an increase in the electron cloud density of benzene in BPA, promoting charge transfer interactions (as electron donors) and becoming vulnerable to ROS attacks. This could explain the superior removal rate to that of p-CR. Remarkably, pre-adsorption hindered p-CR and BPA degradation, as pre-adsorbed pollutants blocked PMS activation sites, delaying ROS generation. In contrast, simultaneous adsorption and catalysis allowed dissolved PMS to rapidly generate abundant ROS surrounding the NG/NH₂-MCM-41, and then rapidly degrade the p-CR and BPA which were adsorbed on NG/NH₂-MCM-41 [102].

$${\mathrm{e}}^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(43)

$${\mathrm{e}}^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to • \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(44)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}-{\mathrm{e}}^{-}\to \mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(45)

4.3. Metallic–Nonmetallic Functionalized Mesoporous Silica Materials

The incorporation of nonmetallic elements on mesoporous silica creates metal-nonmetal sites, which not only prevents metal leaching but also adsorbs pollutants and persulfates [118,119]. Critically, these sites act as electron mediators, accelerating the redox cycling of the metal centers. In the GO and Fe loaded MCM-41 composites (GO-MCM-Fe), GO enhanced Fe stability and prolonged the service life, and the C = C bonds in GO participated in redox reactions with Fe(III) to promote the regeneration of Fe(II). Specifically, the C = C bonds were oxidized to C = O groups, simultaneously reducing Fe(III) to Fe(II). The regenerated Fe(II) subsequently activated PDS to produce SO₄^•− and •OH to obtain a 97.28% removal rate of LVF within 10 min. Notably, the degradation rate constant was 11.78 times higher than that of the GO-MCM-41/PDS system and 1.35 times higher than that of the MCM-Fe/PDS system [120].

The doping of nonmetallic C and N is commonly used in metallic–nonmetallic functionalized mesoporous silica. Yang et al. [84] identified Fe-N₄ and Fe₂-N₂ of Fe-Im-SBA-15 as primary active sites to activate PMS to generate ¹O₂ and O₂^•− for RhB degradation. The RhB removal rate was 97% in 5 min, far surpassing systems lacking Fe, imidazole, or SBA-15. This superior activity arose from the catalytic sites of Fe-imidazole complexes and the homogeneous dispersion induced by the silica support. In addition, high calcination temperature further contributed to RhB removal by promoting the generation of graphitic carbon and active species. Zhang et al. [61] clarified that the coordination of Fe-N and Fe-C in the Fe-C/N-SBA-15 was conductive to the formation of highly stabilized isolated ultra-fine α-Fe₂O₃ nanoparticles, which complexed with TC via -NH₂ groups. This catalyst enhanced both TC adsorption and PMS activation, and obtained a TC removal of more than 90% in 35 min via the produced SO₄^•−, •OH, O₂^•−, and ¹O₂. The doping of C and N was also applied to enhance the catalytic ability for Co-containing mesoporous silica. CoCNx@SBA-15/PMS exhibited a naproxen (NPX) degradation kinetic constant of 0.0877 min⁻¹, more than the Co@SBA-15/PMS (0.0011 min⁻¹) and CNx@SBA-15/PMS (0.0337 min⁻¹) systems. This enhancement resulted from the confinement effect caused by dispersing Co sites bound to pyridinic N and encapsulated CNx layers in the channels of SBA-15. Electrons were transferred from Co(II) to the CNx layers, which facilitated the activation of PMS to generate SO₄^•− and •OH (Equations (46) and (47)). Simultaneously, electron transfer between the C atoms of CNx and PMS resulted in the formation of SO₅^•− (Equation (48)). Subsequently, ¹O₂ was produced through reactions involving SO₅^•− and SO₅^•−/H₂O (Equations (34) and (49)). Collectively, all of SO₄^•−, •OH, and ¹O₂ contributed to the degradation of NPX via hydroxylation and polymerization pathways [83]. Luo et al. [95] designed Co@NC-ZS with short-channel lamellar structures, where Co/CoO nanoparticles and N/C sites were confined within Zr-SBA-15 pores. The activation mechanism is illustrated in Equations (3), (24) and (50)–(55). Co⁰ was able to activate PMS to yield Co(II) and SO₄^•−. Subsequently, SO₄^•− and •OH were formed through the reaction between Co(II) and PMS. Notably, Co⁰ facilitated the regeneration of Co(II). Meanwhile, the non-radical ¹O₂ was generated via the electron transfer from electron-deficient C to PMS. This catalytic system achieved a 96.1% removal of BPA in 40 min. Moreover, BPA removal efficiency first increased and then decreased, with calcination temperature raising from 600 °C to 900 °C, and reached an optimum performance at 700 °C due to the achievement of a maximal graphitic N density and an optimal Co/CoO content. This was because graphitic N with high electronegativity promoted charge transfer, and metallic Co facilitated the SO₄^•− generation through electron donation. Co/N@ZS-SA presented an excellent CIP removal rate (82.52% in 60 min) via SO₄^•−, •OH, O₂^•−, and ¹O₂ pathways [96].

$$\stackrel{\mathrm{e}}{\overbrace{\mathrm{C}\mathrm{o}\mathrm{C}\mathrm{N}\mathrm{x}@\mathrm{S}\mathrm{B}\mathrm{A}-15+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}}}\to • \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(46)

$$\stackrel{\mathrm{e}}{\overbrace{\mathrm{C}\mathrm{o}\mathrm{C}\mathrm{N}\mathrm{x}@\mathrm{S}\mathrm{B}\mathrm{A}-15+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}}}\to \mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(47)

$$\stackrel{\mathrm{e}}{\overbrace{\mathrm{C}\mathrm{o}\mathrm{C}\mathrm{N}\mathrm{x}@\mathrm{S}\mathrm{B}\mathrm{A}-15+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}}}\to \mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(48)

$$\mathrm{S}{\mathrm{O}}_5^{•-}+\mathrm{S}{\mathrm{O}}_5^{•-}\to {}^1\mathrm{O}_2+{\mathrm{S}}_2{\mathrm{O}}_8^{2-}$$

(49)

$${\mathrm{C}\mathrm{o}}^0+2\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{2+}+2\mathrm{S}{\mathrm{O}}_4^{•-}+2\mathrm{O}{\mathrm{H}}^{-}$$

(50)

$${\mathrm{C}\mathrm{o}}^{2+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{3+}+• \mathrm{O}\mathrm{H}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(51)

$${\mathrm{C}\mathrm{o}}^{2+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{3+}+\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(52)

$${\mathrm{C}\mathrm{o}}^{3+}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{2+}+\mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}$$

(53)

$${\mathrm{C}\mathrm{o}}^0+2{\mathrm{C}\mathrm{o}}^{3+}\to 3{\mathrm{C}\mathrm{o}}^{2+}$$

(54)

$$\mathrm{S}{\mathrm{O}}_4^{•-}/•\mathrm{O}\mathrm{H}+\mathrm{p}\mathrm{u}\mathrm{l}\mathrm{l}\mathrm{u}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}\to \mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{s}+{\mathrm{H}}_2\mathrm{O}+\mathrm{C}{\mathrm{O}}_2$$

(55)

The introduction of heteroatomic S further advantages the catalytic performance of metal–nonmetal bi-functionalized mesoporous silica. Comparative studies were conducted on CBZ removal by Fe@SBA-15, Fe@C@SBA-15, Fe@C-N@SBA-15, C-N-S@SBA-15, γ-Fe₂O₃, Fe@C-N-S and Fe@C-N-S@SBA-15 in the presence of PMS [97]. It was found that the FeS component in Fe@C-N-S@SBA-15 as a primary active center for PMS activation provided abundant interfacial Fe(II) sites. The C-N-S component stabilized γ-Fe₂O₃/FeS nanoparticles and promoted the Fe(II)/Fe(III) redox cycle via electron transfer, while γ-Fe₂O₃ synergized with N-doped carbon to form Fe-N-C sites for assisting PMS activation and CBZ adsorption. The SBA-15 framework offered confined pyrolysis spaces to disperse γ-Fe₂O₃, FeS, and C-N-S homogeneously, and its hydrophilicity improved aqueous dispersibility, enabling a nearly 100% CBZ degradation rate within 25 min via the •OH-dominated radical pathway. The generated •OH generation occurred via two primary mechanisms: direct activation of PMS by the intimately coupled FeS/C-N-S composite and activation by released Fe(II) ions from FeS. However, excessive C-N-S content hindered CBZ degradation by blocking accessibility to active sites, despite enhancing CBZ adsorption. Furthermore, the catalyst exhibited strong magnetic property, allowing easy magnetic separation via external magnets for potential reuse.

Table 2. Functionalized mesoporous silica-based persulfate activation for pollutant removal in water.

Catalyst	Persulfate	Pollutant	Reaction Conditions	Dispersion State of Active Component	Degradation Pathway	Kinetic Model	Removal Rate	Ref.
Co-MCM-41	PMS	Orange G	Catalyst = 100 mg, PMS: orange G = 17:1, orange G = 45 mg/L, pH = 6	Co3O4			99% (120 min)	[82]
Co/SBA-15-Cl	PMS	Phenol	Catalyst = 0.2 g/L, PMS = 2 g/L, phenol = 30 mg/L, T = 25 °C	Co3O4	SO4•−	Zero order kinetic model (0.1725 ppm/min)	100% (200 min)	[58]
Co3O4@SBA-15	PMS	SMX	Catalyst = 0.1 g/L, PMS = 5.0 Mm, SMX = 0.04 mM, T = 25 °C	Co3O4	Electron transfer		100% (50 min)	[80]
Co@MCM-41	PMS	AZRS	Catalyst = 0.05 g/L, PMS = 50 mg/L, T = 25 °C	Co3O4	1O2, SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.196 min−1)	100% (30 min)	[91]
Co@SBA-16/ceramic	PMS	LVF	PMS = 0.075 mM, LVF = 10 mg/L, pH = 7.5, flow = 2 mL/min, machine tool height = 2 cm	Co3O4	1O2, O2•−, SO4•− and •OH		78% (180 min)	[71]
QS-CoS-SBA-15	PMS	Phenol	Catalyst = 0.2 g/L, PMS = 2.0 g/L, phenol = 20 mg/L, T = 25 °C	Co-O-Si bonds	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.32 min−1)	100% (10 min)	[92]
Co/SBA-15	PMS	Phenol	Catalyst = 0.1 g/L, PMS: phenol = 4:1, phenol = 50 mg/L, pH = 7, T = 25 °C	Cubic spinel Co3O4 and Co-O-Si species	SO4•−		>98% (120 min)	[70]
Cu -MCM-41	PDS	Orange G	Catalyst = 100 mg, PDS: orange G = 25:1, orange G = 45 mg/L, pH = 6	CuO			80% (120 min)	[82]
Cu-BMS	PDS	MB	Catalyst = 0.5 g/L, PDS = 2.0 g /L, MB = 100 mg/L, pH = 6.5	CuO	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.0641 min−1)	93.5% (60 min)	[112]
CuO/MSS	PDS	BPA	Catalyst = 1.5 g/L, PDS = 100 mM, BPA = 50 mg/L, pH = 7.0, T = 25 °C	CuO	O2•−, SO4•−, •OH and electron transfer	Pseudo-first-order kinetic model (k = 0.14 min−1)	92% (45 min)	[100]
Cu-SBA-15	PMS	TC	Catalyst = 20 mg, PMS = 25 mg, TC = 6 mg/L	CuO and single-atom Cu	1O2, O2•−, SO4•− and •OH		90% (60 min)	[60]
MnOx/SBA-15	PMS	BPB	Catalyst = 0.5 g/L, PMS = 5 mM, BPB = 80 mL, pH = 6.5 ± 0.2, T = 25 °C	MnOx (MnO, Mn3O4, Mn2O3 and MnO2 )	SO4•− and •OH		98.4% (180 min)	[81]
Mn/MSS	PMS	MO	Catalyst = 40 mg, PMS = 1 mM, MO = 100 ppm, pH = 7, T = 25 °C	MnOx (Mn(II)/Mn(III) coexistence)			90% (8 min)	[59]
Mn/asSBA-15	PDS	SCP	Catalyst = 0.2 g/L, PDS = 2.0 g/L, SCP = 20 mg/L, T = 25 °C	Mn2O3	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.017 min−1)	100% (330 min)	[90]
Fe3O4@SBA-15	PDS	CBZ	Catalyst = 0.5 g/L, PDS = 300 mg/L, CBZ = 10 mg/L, pH = 3.0, T = 25 °C	Fe3O4	SO4•−, O2•−and •OH	Pseudo-first-order kinetic model (k = 0.0023 min−1)	100% (30 min)	[105]
Fe-MSS	PDS	MB	Catalyst = 1 g/L, PDS = 1 g/L, MB = 50 mg/L, pH = 5	Fe3O4	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.1402 min−1)	90% (60 min)	[115]
Fe@void@mSiO2	PMS	TC	Catalyst = 0.1 g/L, PMS = 0.18 g/L, TC = 10 mg/L, T = 20 °C	FeO	SO4•− and •OH		100% (40 min)	[62]
CF/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS: RhB = 20:1, RhB = 5 mg/L, T = 25 °C	CoFe2O4	SO4•−	First-order kinetic model (k = 0.032 min−1)	98% (120 min)	[72]
CFMCM-41	PMS	CIP	Catalyst = 0.1 g/L, PMS = 0.37 g/L, CIP = 5 mg/L, pH = 7.0	CoFe2O4	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.17 min−1)	>90% (60 min)	[106]
CoFe2O4-MCM-41	PMS	SMX	Catalyst = 0.2 g/L, PMS = 0.15 mM, SMX = 10 mg/L, pH = 7.0, T = 25 ± 1 °C	CoFe2O4	SO4•−, •OH, O2•− and electron transfer	Second-order kinetic model (k = 0.10244 min−1·mg−1·L)	96.82% (30 min )	[93]
CFS	PMS	Acid orange II	Catalyst = 0.2 g/L, PMS = 1 g/L, acid orange II = 20 mg/L, T = 30 °C	CoFe2O4	SO4•− and •OH		>90% (35 min)	[117]
FeCo-MCM-41	PMS	MO	Catalyst = 0.2 g/L, PMS = 0.075 mM, MO = 25 mg/L, pH = 5.6, T = 25 °C	Si-O-Fe, atomically dispersed Co(II)	1O2 and SO4•−		~95% (15 min)	[63]
γ-Fe2O3@2Co-MCM-41	PMS	Acid orange II	Catalyst = 0.2 g/L, PMS = 1.2 mM, acid orange II = 0.2 mM, T = 25 °C	Co(II) and γ-Fe2O3	SO4•− and 1O2	Pseudo-first-order kinetic model (k = 0.0778 min−1)	99% (40 min)	[65]
CoMn/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS = 0.3 mmol/L, RhB = 50 mg/L, pH = 4.2, T = 30 °C	MnCo2O4	1O2, SO4•− and •OH		99% (45 min)	[73]
Mn-Co-MCM-41	PMS	RhB	Catalyst = 5 mg, PMS = 0.2 mmol/L, RhB= 5 mg/L	Mn-Co bimetallic oxide nanoparticles	SO4•−, •OH and O2•−		98% (20 min)	[64]
Cu0.76Co2.24O4/SBA-15	PMS	SPD	Catalyst = 1.0 g/L, PMS: SPD = 20:1, SPD = 50 μmol/L, pH = 7	Cu0.76Co2.24O4 spinel structure	1O2, SO4•−, •OH and O2•−	Pseudo-first-order kinetic model (k = 0.058 min−1)	99.2% (90 min)	[74]
AgCo/SBA-15	PMS	UDMH	Catalyst = 0.5 g/L, PMS = 33.3 mM, UDMH = 100 mg/L, pH = 9.0, T = 30 °C	Co3O4 and Ag nanoparticles	1O2, SO4•− and •OH	First-order kinetic model (k = 0.1602 min−1)	100% (15 min)	[78]
CoMg/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS: RhB = 10:1, RhB = 5 mg/L, T = 25 °C	Co3O4 and highly dispersed MgO	SO4•−	First-order kinetic model (k = 1.065 min−1)	100% (5 min)	[79]
LaCoO3/SBA-15	PMS	ATR	Catalyst = 0.6 g/L, PMS = 4.29 mM, ATR = 10 mg/L, pH = 8.2, T = 25 °C	LaCoO3	SO4•−, •OH, 1O2 and O2•−		~100% (6 min)	[75]
Cu0.2Ni0.8O/SBA-15	PMS	SCP	Catalyst = 1.5 g/L, PMS: SCP = 20:1, SCP = 50 μmol/L, pH = 7.0	Cu0.2Ni0.8O	1O2, O2•−, SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.0557 min−1)	99.27% (90 min)	[76]
Cu-Al/MSS	PDS	BPA	Catalyst = 0.5 g/L, PDS = 100 mmol/L, BPA = 50 mg/L, pH = 7, T = 25 °C	CuO and Al2O3	1O2, SO4•−, •OH, O2•− and electron transfer	First-order kinetic model (k = 0.21 min−1)	90% (30 min)	[101]
Sr2FeO4 /SBA-15	PMS	SPD	Catalyst = 0.07 g/L, PMS: SPD = 40:1, SPD = 50 μmol/L, pH = 7.0	Sr2FeO4 spinel nanoparticles	SO4•−and •OH	First-order kinetic model (k = 0.0548 min−1)	99.0% (90 min)	[77]
g-C3N4/MCM-41	PMS	AO7	Catalyst = 1.0 g/L, PMS = 0.188 g/L, pollutants = 5 mg/L, T = 27 °C	g-C3N4	SO4•−, •OH and electron transfer	First-order kinetic model (k = 0.113 min−1)	96.8% (30 min)	[94]
NG/NH2-MCM-41	PMS	p-CR, BPA	Catalyst = 0.2 g/L, PMS = 0.2 g/L, pollutants = 5 mg/L, pH = 6.0, T = 25 °C	NG layers	1O2, SO4•− and •OH		p-CR: 95%, BPA: 100% (700 min)	[102]
GO-MCM-Fe	PDS	LVF	Catalyst = 0.5 mg/L, PMS = 0.2 mg/L, LVF = 100 mg/L, pH = 4.3, T = 20 °C	Fe oxide nanoparticles and GO	SO4•− and •OH	First-order kinetic model (k = 0.35405 min−1)	97.28% (10 min)	[120]
Fe-Im-SBA-15	PMS	RhB	Catalyst = 0.12 g/L, PMS = 2.4 mM, RhB = 50 mg/L, pH = 7	Fe-imidazole complexes and graphitic C	1O2 and O2•−		97.0% (5 min)	[84]
Fe-C/N-SBA-15	PMS	TC	Catalyst = 15 mg, PMS = 20 mg, TC = 10 mg/L	Fe-N, Fe-C bonds and Fe2O3	SO4•−, •OH, O2•− and 1O2		>90% (35 min)	[61]
CoCNx@SBA-15	PMS	NPX	Catalyst = 0.0375 g/L, PMS = 2.5 mM, NPX = 0.043 mM, T = 25 °C	CoO and Co-N coordination	1O2, SO4•− and •OH	First-order kinetic model (k = 0.0877 min−1)	100% (55 min)	[83]
Co@NC-ZS	PMS	BPA	Catalyst = 0.1 g/L, PMS = 0.3 g/L, BPA = 20 mg/L, pH = 6.2, T = 25 °C	Co/CoO nanoparticles and N-C layer	SO4•−, •OH and 1O2	Pseudo-first-order kinetic model (k = 0.0778 min−1)	96.1% (30 min)	[95]
Co/N@ZS-SA	PMS	CIP	Catalyst = 0.6 g/L, PMS = 0.50 g/L, CIP = 20 mg/L, pH = 6.7	Co nanoparticles	1O2, SO4•−, •OH and O2•−	Pseudo-first-order kinetic model (k = 0.0378 min−1)	82.52% (60 min)	[96]
Fe@C-N-S@SBA-15	PMS	CBZ	Catalyst = 0.6 g/L, PMS = 1.0 mM, CBZ= 10 mg/L	FeS, γ-Fe2O3 nanoclusters and N/S-doped C	•OH		~100% (25 min)	[97]

Table 2. Functionalized mesoporous silica-based persulfate activation for pollutant removal in water.

Catalyst	Persulfate	Pollutant	Reaction Conditions	Dispersion State of Active Component	Degradation Pathway	Kinetic Model	Removal Rate	Ref.
Co-MCM-41	PMS	Orange G	Catalyst = 100 mg, PMS: orange G = 17:1, orange G = 45 mg/L, pH = 6	Co3O4			99% (120 min)	[82]
Co/SBA-15-Cl	PMS	Phenol	Catalyst = 0.2 g/L, PMS = 2 g/L, phenol = 30 mg/L, T = 25 °C	Co3O4	SO4•−	Zero order kinetic model (0.1725 ppm/min)	100% (200 min)	[58]
Co3O4@SBA-15	PMS	SMX	Catalyst = 0.1 g/L, PMS = 5.0 Mm, SMX = 0.04 mM, T = 25 °C	Co3O4	Electron transfer		100% (50 min)	[80]
Co@MCM-41	PMS	AZRS	Catalyst = 0.05 g/L, PMS = 50 mg/L, T = 25 °C	Co3O4	1O2, SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.196 min−1)	100% (30 min)	[91]
Co@SBA-16/ceramic	PMS	LVF	PMS = 0.075 mM, LVF = 10 mg/L, pH = 7.5, flow = 2 mL/min, machine tool height = 2 cm	Co3O4	1O2, O2•−, SO4•− and •OH		78% (180 min)	[71]
QS-CoS-SBA-15	PMS	Phenol	Catalyst = 0.2 g/L, PMS = 2.0 g/L, phenol = 20 mg/L, T = 25 °C	Co-O-Si bonds	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.32 min−1)	100% (10 min)	[92]
Co/SBA-15	PMS	Phenol	Catalyst = 0.1 g/L, PMS: phenol = 4:1, phenol = 50 mg/L, pH = 7, T = 25 °C	Cubic spinel Co3O4 and Co-O-Si species	SO4•−		>98% (120 min)	[70]
Cu -MCM-41	PDS	Orange G	Catalyst = 100 mg, PDS: orange G = 25:1, orange G = 45 mg/L, pH = 6	CuO			80% (120 min)	[82]
Cu-BMS	PDS	MB	Catalyst = 0.5 g/L, PDS = 2.0 g /L, MB = 100 mg/L, pH = 6.5	CuO	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.0641 min−1)	93.5% (60 min)	[112]
CuO/MSS	PDS	BPA	Catalyst = 1.5 g/L, PDS = 100 mM, BPA = 50 mg/L, pH = 7.0, T = 25 °C	CuO	O2•−, SO4•−, •OH and electron transfer	Pseudo-first-order kinetic model (k = 0.14 min−1)	92% (45 min)	[100]
Cu-SBA-15	PMS	TC	Catalyst = 20 mg, PMS = 25 mg, TC = 6 mg/L	CuO and single-atom Cu	1O2, O2•−, SO4•− and •OH		90% (60 min)	[60]
MnOx/SBA-15	PMS	BPB	Catalyst = 0.5 g/L, PMS = 5 mM, BPB = 80 mL, pH = 6.5 ± 0.2, T = 25 °C	MnOx (MnO, Mn3O4, Mn2O3 and MnO2 )	SO4•− and •OH		98.4% (180 min)	[81]
Mn/MSS	PMS	MO	Catalyst = 40 mg, PMS = 1 mM, MO = 100 ppm, pH = 7, T = 25 °C	MnOx (Mn(II)/Mn(III) coexistence)			90% (8 min)	[59]
Mn/asSBA-15	PDS	SCP	Catalyst = 0.2 g/L, PDS = 2.0 g/L, SCP = 20 mg/L, T = 25 °C	Mn2O3	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.017 min−1)	100% (330 min)	[90]
Fe3O4@SBA-15	PDS	CBZ	Catalyst = 0.5 g/L, PDS = 300 mg/L, CBZ = 10 mg/L, pH = 3.0, T = 25 °C	Fe3O4	SO4•−, O2•−and •OH	Pseudo-first-order kinetic model (k = 0.0023 min−1)	100% (30 min)	[105]
Fe-MSS	PDS	MB	Catalyst = 1 g/L, PDS = 1 g/L, MB = 50 mg/L, pH = 5	Fe3O4	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.1402 min−1)	90% (60 min)	[115]
Fe@void@mSiO2	PMS	TC	Catalyst = 0.1 g/L, PMS = 0.18 g/L, TC = 10 mg/L, T = 20 °C	FeO	SO4•− and •OH		100% (40 min)	[62]
CF/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS: RhB = 20:1, RhB = 5 mg/L, T = 25 °C	CoFe2O4	SO4•−	First-order kinetic model (k = 0.032 min−1)	98% (120 min)	[72]
CFMCM-41	PMS	CIP	Catalyst = 0.1 g/L, PMS = 0.37 g/L, CIP = 5 mg/L, pH = 7.0	CoFe2O4	SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.17 min−1)	>90% (60 min)	[106]
CoFe2O4-MCM-41	PMS	SMX	Catalyst = 0.2 g/L, PMS = 0.15 mM, SMX = 10 mg/L, pH = 7.0, T = 25 ± 1 °C	CoFe2O4	SO4•−, •OH, O2•− and electron transfer	Second-order kinetic model (k = 0.10244 min−1·mg−1·L)	96.82% (30 min )	[93]
CFS	PMS	Acid orange II	Catalyst = 0.2 g/L, PMS = 1 g/L, acid orange II = 20 mg/L, T = 30 °C	CoFe2O4	SO4•− and •OH		>90% (35 min)	[117]
FeCo-MCM-41	PMS	MO	Catalyst = 0.2 g/L, PMS = 0.075 mM, MO = 25 mg/L, pH = 5.6, T = 25 °C	Si-O-Fe, atomically dispersed Co(II)	1O2 and SO4•−		~95% (15 min)	[63]
γ-Fe2O3@2Co-MCM-41	PMS	Acid orange II	Catalyst = 0.2 g/L, PMS = 1.2 mM, acid orange II = 0.2 mM, T = 25 °C	Co(II) and γ-Fe2O3	SO4•− and 1O2	Pseudo-first-order kinetic model (k = 0.0778 min−1)	99% (40 min)	[65]
CoMn/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS = 0.3 mmol/L, RhB = 50 mg/L, pH = 4.2, T = 30 °C	MnCo2O4	1O2, SO4•− and •OH		99% (45 min)	[73]
Mn-Co-MCM-41	PMS	RhB	Catalyst = 5 mg, PMS = 0.2 mmol/L, RhB= 5 mg/L	Mn-Co bimetallic oxide nanoparticles	SO4•−, •OH and O2•−		98% (20 min)	[64]
Cu0.76Co2.24O4/SBA-15	PMS	SPD	Catalyst = 1.0 g/L, PMS: SPD = 20:1, SPD = 50 μmol/L, pH = 7	Cu0.76Co2.24O4 spinel structure	1O2, SO4•−, •OH and O2•−	Pseudo-first-order kinetic model (k = 0.058 min−1)	99.2% (90 min)	[74]
AgCo/SBA-15	PMS	UDMH	Catalyst = 0.5 g/L, PMS = 33.3 mM, UDMH = 100 mg/L, pH = 9.0, T = 30 °C	Co3O4 and Ag nanoparticles	1O2, SO4•− and •OH	First-order kinetic model (k = 0.1602 min−1)	100% (15 min)	[78]
CoMg/SBA-15	PMS	RhB	Catalyst = 0.1 g/L, PMS: RhB = 10:1, RhB = 5 mg/L, T = 25 °C	Co3O4 and highly dispersed MgO	SO4•−	First-order kinetic model (k = 1.065 min−1)	100% (5 min)	[79]
LaCoO3/SBA-15	PMS	ATR	Catalyst = 0.6 g/L, PMS = 4.29 mM, ATR = 10 mg/L, pH = 8.2, T = 25 °C	LaCoO3	SO4•−, •OH, 1O2 and O2•−		~100% (6 min)	[75]
Cu0.2Ni0.8O/SBA-15	PMS	SCP	Catalyst = 1.5 g/L, PMS: SCP = 20:1, SCP = 50 μmol/L, pH = 7.0	Cu0.2Ni0.8O	1O2, O2•−, SO4•− and •OH	Pseudo-first-order kinetic model (k = 0.0557 min−1)	99.27% (90 min)	[76]
Cu-Al/MSS	PDS	BPA	Catalyst = 0.5 g/L, PDS = 100 mmol/L, BPA = 50 mg/L, pH = 7, T = 25 °C	CuO and Al2O3	1O2, SO4•−, •OH, O2•− and electron transfer	First-order kinetic model (k = 0.21 min−1)	90% (30 min)	[101]
Sr2FeO4 /SBA-15	PMS	SPD	Catalyst = 0.07 g/L, PMS: SPD = 40:1, SPD = 50 μmol/L, pH = 7.0	Sr2FeO4 spinel nanoparticles	SO4•−and •OH	First-order kinetic model (k = 0.0548 min−1)	99.0% (90 min)	[77]
g-C3N4/MCM-41	PMS	AO7	Catalyst = 1.0 g/L, PMS = 0.188 g/L, pollutants = 5 mg/L, T = 27 °C	g-C3N4	SO4•−, •OH and electron transfer	First-order kinetic model (k = 0.113 min−1)	96.8% (30 min)	[94]
NG/NH2-MCM-41	PMS	p-CR, BPA	Catalyst = 0.2 g/L, PMS = 0.2 g/L, pollutants = 5 mg/L, pH = 6.0, T = 25 °C	NG layers	1O2, SO4•− and •OH		p-CR: 95%, BPA: 100% (700 min)	[102]
GO-MCM-Fe	PDS	LVF	Catalyst = 0.5 mg/L, PMS = 0.2 mg/L, LVF = 100 mg/L, pH = 4.3, T = 20 °C	Fe oxide nanoparticles and GO	SO4•− and •OH	First-order kinetic model (k = 0.35405 min−1)	97.28% (10 min)	[120]
Fe-Im-SBA-15	PMS	RhB	Catalyst = 0.12 g/L, PMS = 2.4 mM, RhB = 50 mg/L, pH = 7	Fe-imidazole complexes and graphitic C	1O2 and O2•−		97.0% (5 min)	[84]
Fe-C/N-SBA-15	PMS	TC	Catalyst = 15 mg, PMS = 20 mg, TC = 10 mg/L	Fe-N, Fe-C bonds and Fe2O3	SO4•−, •OH, O2•− and 1O2		>90% (35 min)	[61]
CoCNx@SBA-15	PMS	NPX	Catalyst = 0.0375 g/L, PMS = 2.5 mM, NPX = 0.043 mM, T = 25 °C	CoO and Co-N coordination	1O2, SO4•− and •OH	First-order kinetic model (k = 0.0877 min−1)	100% (55 min)	[83]
Co@NC-ZS	PMS	BPA	Catalyst = 0.1 g/L, PMS = 0.3 g/L, BPA = 20 mg/L, pH = 6.2, T = 25 °C	Co/CoO nanoparticles and N-C layer	SO4•−, •OH and 1O2	Pseudo-first-order kinetic model (k = 0.0778 min−1)	96.1% (30 min)	[95]
Co/N@ZS-SA	PMS	CIP	Catalyst = 0.6 g/L, PMS = 0.50 g/L, CIP = 20 mg/L, pH = 6.7	Co nanoparticles	1O2, SO4•−, •OH and O2•−	Pseudo-first-order kinetic model (k = 0.0378 min−1)	82.52% (60 min)	[96]
Fe@C-N-S@SBA-15	PMS	CBZ	Catalyst = 0.6 g/L, PMS = 1.0 mM, CBZ= 10 mg/L	FeS, γ-Fe2O3 nanoclusters and N/S-doped C	•OH		~100% (25 min)	[97]

5. Influence of Reaction Conditions

5.1. Catalyst Dosage

The dosage of catalysts directly governs the number of active sites in heterogeneous reactions. Most studies have indicated a significant enhancement in pollutant degradation efficiency, with an increase of material dosage, due to the more available active sites for persulfate activation [58,61,91,96]. Yu et al. [76] have reported that increasing the dosage of Cu_0.2Ni_0.8O/SBA-15 from 0.25 g/L to 2.0 g/L improved the SCP removal rate from 77.70% to 99.92% in the presence of PMS, attributing to the enhanced electron transfer efficiency from the catalyst to PMS. Sun et al. [63] have suggested that the MO removal improved from 76% to 94% by PMS with the FeCo-MCM-41 dosage raising from 0.05 to 0.3 g/L, as the catalyst provided additional sites for MO adsorption and PMS activation. However, a number of studies indicated no further significant improvement in the pollutant elimination beyond an optimal catalyst dosage. Luo et al. [95] observed that elevating Co@NC-ZS dosage from 0.05 g/L to 0.10 g/L improved BPA removal from 88.4% to 96.1%, but further increasing the dosage to 0.15 g/L only marginally enhanced BPA removal rate to 97%, due to the fixed PMS concentration. The ATR elimination raised from 84.5% to 100%, as LaCoO₃/SBA-15 increased from 0.2 to 0.6 g/L. However, only a minimal reduction was observed in degradation time, from 6 min to 5 min, due to the quenching effects, slight aggregation of catalysts, or inefficient utilization of excess active sites [75]. Similar studies were also investigated by He et al. [74], Gao et al. [93] and Yang et al. [81].

Conversely, some studies have suggested that pollutants degradation efficiency decreased beyond the optimal catalyst dosage condition. Liang et al. [100] found that a further increase of CuO/MSS from 1.5 g/L to 2 g/L decreased the BPA elimination from 92% to 88.6%. This was ascribed to the limited available active sites, where an excess of catalyst decreased the probability of the target pollutants occupying adjacent sites to reactive species. Similar behavior has been exhibited in the Cu-Al/MSS/PDS system for BPA removal [101]. Additionally, Liang et al. [115] found that further increasing Fe-MSS dosage from 1 g/L to 2 g/L reduced the MB removal efficiency from 90% to 85%, which was attributed to the excessive release of soluble Fe(II). The surplus Fe(II) reacted with SO₄⁻·, reducing ROSs and consequently lowering the degradation efficiency.

Overall, excessive catalyst dosage is capable of increasing operational costs, which hinders practical applications. Therefore, selecting an optimal catalyst dosage is crucial in experimental design.

5.2. Persulfate Dosage

Persulfate concentration exhibited a similar trend to catalyst dose in enhancing contaminant elimination, where an increase of persulfate concentration promoted reaction efficiency through enhanced ROS generation [64,78,80,91]. However, several studies have demonstrated that excessive persulfate dosage often yielded diminishing improvements or even reduced removal efficiency beyond optimal concentrations. Luo et al. [95] presented a slight increase of BPA decomposition from 96.1% to 98.2% by Co@NC-ZS, with PMS concentration raising from 0.3 to 0.35 g/L, owing to the limited active sites of the catalyst. These results have also been elucidated by Xu et al. [73] and Sun et al. [63]. Gao et al. [93] noted that ROS scavenging by excess PMS (Equations (49) and (56)–(58)) was responsible for this phenomenon, something which Zhao et al. [77] and Sajjadi et al. agreed upon [112]. Afzal et al. [75] indicated that both of the above two situations led to a decline in ATR removal by the LaCoO₃/SBA-15/PMS system. Wang et al. [101] implied that an excessive PMS resulted in a decrease in BPA removal efficiency by the Cu-Al/MSS-driven PDS system, due to radical quenching (Equation (59)). A similar trend was also presented by Liang et al. [115], Choong et al. [106] and Liang et al. [100].

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+\mathrm{S}{\mathrm{O}}_4^{•-}\to \mathrm{S}{\mathrm{O}}_5^{•-}+{\mathrm{H}}^{+}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(56)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+• \mathrm{O}\mathrm{H}\to {\mathrm{H}}_2\mathrm{O}+\mathrm{S}{\mathrm{O}}_5^{•-}$$

(57)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{S}{\mathrm{O}}_4^{•-}\to {\mathrm{S}}_2{\mathrm{O}}_8^{2-}$$

(58)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{S}}_2{\mathrm{O}}_8^{2-}\to {\mathrm{S}}_2{\mathrm{O}}_8^{•-}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(59)

5.3. Initial Pollutant Concentration

A well-documented negative correlation exists between initial contaminant concentration and its removal rate in functionalized mesoporous silica/persulfate systems. This phenomenon was primarily attributed to the following two reasons: (1) Insufficient generated ROSs to degrade high pollutant concentrations of pollutants, and (2) Competitive adsorption of high-concentration pollutants and their intermediates on catalyst surfaces, impeding degradation performance. A drastic reduction observation in SMX degradation rate constant from 1.19261 to 0.00512 min⁻¹·mg⁻¹·L by the Co₂FeO₄-MCM-41/PMS system, when SMX concentration increased from 5 to 20 mg/L. This was attributed to the inadequate ROSs for high contaminant concentration elimination [93]. Sajjadi et al. [112] established that both ROS inadequacy and MB-induced active site blockage were responsible for the dropped removal of MB by the co-existence of PDS and Cu-BMS with increasing MB concentration. He et al. [74] revealed the inhibition of SPD concentration was due not only to the adsorption/desorption behavior of intermediate products, but also to the impediment of electron transfer from PMS by the produced electrophilic intermediates (e.g., -NH₂, C≡N). Comparable concentration-dependent inhibition was documented for CIP degradation by Co/N@ZS-SA [96], ATR removal by LaCoO₃/SBA-15 [75], SCP elimination by Cu_0.2Ni_0.8O/SBA-15 [76], UDMH decomposition by AgCo/SBA-15 [78], and SPD degradation by Sr₂FeO₄/SBA-15 [77] in presence of PMS.

Interestingly, Hu et al. [70] showed a positive relationship between initial phenol concentration and phenol removal efficiency, as well as Co leaching, in the Co/SBA-15/PMS system. They attributed the reduced leaching at lower phenol concentrations to diminished adsorption of carboxylic acid by-products on the material, thereby mitigating localized acidification-induced Co dissolution.

5.4. Temperature

Elevating temperature generally enhances the catalytic efficiency, such as in the cases of MB removal by the Fe-MSS/PDS system [115], BPA elimination by the CuO/MSS/PDS system [100] and the Cu-Al/MSS/PDS system [101], MO decomposition by the Cu/MCM-41-NH₂-PDS system [85] and FeCo-MCM-41/PMS system [63], MB degradation by the Cu-BMS/PDS system [112], phenol removal by the Co/SBA-15-Cl/PMS system [58], AZRS decomposition by the Co@MCM-41/PMS system [91], NPX elimination by the CoCNx@SBA-15/PMS system [83], and acid orange II removal by the CFS/PMS system [117].

For instance, temperature increase (15 °C to 45 °C) enhanced the BPA degradation efficiency (82.5% to approximately 92.5%) by the CuO/MSS-activated PDS system [100]. The MO elimination by PDS using Cu/MCM-41-NH₂ as a heterogeneous catalyst accelerated significantly, shifting the required time from 80 min at 30 °C to 40 min at 50 °C [85]. Wang et al. [101] attributed the enhanced BPA removal by the Cu-Al/MSS/PDS system at higher temperatures to the endothermic nature of the reaction, which favored PDS activation and accelerated the interaction between BPA and radical species. It is worth noting that a reduced p-CR removal was observed by PMS in the presence of NG/NH₂-MCM-41/PMS with increasing temperatures, which resulted from the exothermic adsorption of p-CR on the material [102]. In contrast, a minor negative impact on BPA removal was presented at higher temperature by the NG/NH₂-MCM-41/PMS system due to the counteraction of the increased radical activity and decreased adsorption capacity.

5.5. Initial pH

The initial pH in the solution is of primary importance for the functionalized mesoporous silica/persulfate system in pollutant degradation. It critically influences the surface charge of the catalyst, the speciation of the target pollutant and persulfate, as well as the generation of ROS. The reported effects of the initial solution pH on pollutant degradation by the functionalized mesoporous silica-activated PMS system can be roughly summarized into three trends. The first trend is that the catalytic performance of functionalized mesoporous silica initially increased and then decreased with the increase of initial pH. Both excessive acidity and alkalinity inhibited the pollutant removal rate or degradation kinetics constant [71,93,106]. This is mainly because H⁺ tended to form hydrogen bonds with the O-O group of PMS in acidic solutions. Additionally, PMS existed as HSO₅⁻ to generate H₂SO₅ (Equation (60)) [121], which hindered PMS activation. Additionally, the abundant H⁺ also quenched SO₄^•− and •OH (Equations (61) and (62)) [122]. As a result, the pollutant degradation efficiency decreased under acidic conditions. When the pH of the solution increased to an alkaline level (especially pH > 9), PMS primarily existed in the form of SO₅²⁻, which exhibited low catalytic activity [95]. Moreover, the residual HSO₅⁻ was able to react with OH- to regenerate SO₄²⁻ (Equation (63)), and the low redox •OH (E₀ = 1.8–2.7 V) was formed by the reaction between OH⁻ and SO₄^•−, with a high redox potential (E₀ = 2.5–3.1 V) (Equation (24)) [123]. When the functionalized mesoporous silica surface carried a negative charge, electrostatic repulsion occurred between the material with the negatively charged PMS, hindering PMS adsorption and activation [78]. Similarly, if the pollutant also existed in a negatively charged form, electrostatic repulsion between the material and the pollutant further impeded the pollutant adsorption and degradation [74]. The Co complexes/precipitates would be produced in the Co-modified mesoporous silica in alkaline solution, which is unfavorable for persulfate activation [71,75,93]. All of the above contributed to a decline in the pollutant removal by the functionalized mesoporous silica/persulfate system under alkaline conditions. Although a number of studies reported a reduction in pollutant removal rate or degradation kinetics under acidic or alkaline conditions, the extent of reduction was relatively small, indicating a high removal efficiency over a wide initial pH range. Examples included the CuO/MSS/PDS [100], Cu-Al/MSS/PDS [101] and Co@NC-ZS/PMS [95] systems for BPA removal, Co@MCM-41/PMS system for AZRS elimination [91], Fe-MSS/PDS system for MB removal [115], Cu-SBA-15/PMS system for TC removal [60], LaCoO₃/SBA-15/PMS system for ATR decomposition [75], Cu_0.76Co_2.24O₄/SBA-15/PMS system for SPD removal [74], AgCo/SBA-15/PMS system for UDMH decomposition [78], FeCo-MCM-41/PMS system for MO elimination [63], CFMCM-41/PMS system for CIP decomposition [106], Cu_0.2Ni_0.8O/SBA-15/PMS [76] and Sr₂FeO₄/SBA-15/PMS [77] for SCP and SPD elimination, respectively, and NG/NH₂-MCM-41/PMS for p-CR and BPA decomposition [102].

$$\mathrm{H}\mathrm{S}{{\mathrm{O}}_5}^{•-}+{\mathrm{H}}^{+}\to {\mathrm{H}}_2\mathrm{S}{\mathrm{O}}_5$$

(60)

$$\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\to {\mathrm{H}}_2\mathrm{O}$$

(61)

$$• \mathrm{O}\mathrm{H}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\to \mathrm{H}\mathrm{S}{\mathrm{O}}_4^{•-}$$

(62)

$$\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+2\mathrm{O}{\mathrm{H}}^{-}\to 2\mathrm{S}{\mathrm{O}}_4^{2-}+2{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}}_2$$

(63)

The second trend was that the optimal pollutant removal rate was presented in the acidic solution, and that degradation efficiency declined under alkaline conditions. Wang et al. [97] observed that the CBZ removal efficiency by the Fe@C-N-S@SBA-15-driven PMS system decreased from approximately 100% to about 37% as the pH increased from 3.3 to 9.0. This was largely due to the self-quenching of PMS and the shortened lifetime of Fe(II) in the alkaline solution, as Fe(II) acted as PMS activators. The Cu-BMS/PDS system generated more SO₄^•− under the acidic condition during the PMS oxidation process (Equations (3) and (64)–(66)), but high concentrations of OH⁻ quenched SO₄^•− (Equation (24)) [112]. Consequently, the removal rate of MB by the Cu-BMS/PDS system gradually decreased with increasing pH values. Yang et al. [81] found that the removal efficiency of BPB by MnO_x/SBA-15 reached the maximum at an initial pH of 2.93 and slightly declined in the pH range of 4.3–9.95. When the pH further increased to 11.43, there was a significant drop in the BPB elimination. This was attributed to pH-induced changes in the quantity of ROS, namely the weakened signal intensity of •OH and the maintained signal strength of SO₄^•−, evidenced by EPR. This is due to the Lewis acid sites in MnO_x/SBA-15, which absorbed large amounts of OH⁻ ions from the solution and led to a reduction in the activation capacity of PMS, subsequently lowering BPB removal efficiency. However, Huang et al. [105] discovered that the solution pH affected both the adsorption and radical generation processes in the Fe₃O₄@SBA-15/PDS system for CBZ removal. As the pKa of CBZ was 13.90, CBZ existed as a neutral molecule in the pH range of 3–11 and was adsorbed to Fe₃O₄@SBA-15 via hydrogen bonding under acidic conditions. Additionally, Fe₃O₄@SBA-15 was positively charged at pH < 4 and adsorbed negatively charged PDS via electrostatic interaction. Therefore, Fe₃O₄@SBA-15 adsorbed CBZ and S₂O₈²⁻ simultaneously at pH = 3. For degradation, SO₄^•− was the dominant radical for pollutant degradation in acidic solutions in general, whereas SO₄^•− tended to convert to •OH (Equation (24)) in the alkaline solution. As a result, the Fe₃O₄@SBA-15/PDS system achieved nearly a 100% CBZ removal rate at pH = 3.0, but dropped to below 8% at pH = 9 and 11 based on the SO₄^•− dominated pathway.

$${\mathrm{S}}_2{\mathrm{O}}_8^{2-}+{\mathrm{H}}^{+}\to \mathrm{H}{\mathrm{S}}_2{\mathrm{O}}_8^{-}$$

(64)

$$\mathrm{H}{\mathrm{S}}_2{\mathrm{O}}_8^{-}\to \mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{H}}^{+}$$

(65)

$$\mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{O}{\mathrm{H}}^{•}+{\mathrm{H}}^{+}\to \mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{O}$$

(66)

The third trend was that the pollutant degradation efficiency increased with the increasing initial pH of the solution. TC existed as TCH⁻ in the pH range of 7.68–9.70 and as TC²⁻ at pH > 9.70. TCH⁻ and TC²⁻ possessed higher electron density and were preferentially attacked by the radicals generated in the Fe-C/N-SBA-15/PMS system. Additionally, PMS was more prone to producing radicals under alkaline conditions, such as the conversion of SO₄^•− to •OH via reaction with OH⁻ (Equation (24)). Consequently, the TC removal rate at pH =11 was much better than that at pH = 9 and significantly higher than that at pH = 1 [61]. However, Hou et al. [83] have implied that the NPX degradation rate was 92.6% under the acidic condition (pH = 2.7), attributing this to the hydrogen bonding between H⁺ and HSO₅⁻, which reduced PMS activation efficiency. The NPX removal rate reached 100% in the alkaline solutions, primarily due to the significant contribution of PMS self-activation at pH = 11.5.

5.6. Co-Existing Anions

Actual water bodies contain various ions, and they exhibit distinct effects on the target pollutant elimination in the persulfate-based advanced oxidation process. Therefore, this review illustrates the influences of reported anions (Cl⁻, HCO₃⁻, NO₃⁻, H₂PO₄⁻, HPO₄²⁻, SO₄²⁻, CO₃²⁻) on the catalytic performance of functionalized mesoporous silicas comprehensively.

As one of the most common anions in aquatic environments, Cl⁻ frequently demonstrated adverse effects on pollutant removal in functionalized mesoporous silica-activated persulfate systems. Luo et al. [95] reported that the addition of 5 mM Cl⁻ decreased BPA removal from 96.1% to 89.6% in the Co@NC-ZS/PMS system, because it suppressed the SO₄^•− formation and contributed to the generation of less reactive chlorine species like Cl^• and HClO^•− (Equations (67) and (68)). Huang et al. [78] observed that 10 mM Cl⁻ slowed the UDMH degradation rate constant from 0.2840 min⁻¹ to 0.1797 min⁻¹ by AgCo/SBA-15 during the catalytic PMS oxidation, as Cl⁻ quenched SO₄⁻· and ·OH and consumed HSO₅⁻ (Equations (67)–(72)). Similarly, co-existing Cl⁻ (3 mM) caused a decrease from 100% to 87% on CBZ removal in the Fe₃O₄@SBA-15/PDS system within 30 min [105]. When Cl⁻ concentration increased from 0 to 5 mM, the CIP degradation by Co/N@ZS-SA with PMS then declined from 82.52% to 78.23%. With Cl⁻ concentration further raising to 20 mM, CIP degradation reduced to 51.67%. This nonlinear response was predominantly attributed to the progressive consumption of ROS by Cl⁻ (Equation (67)), and the low concentration of Cl⁻ induced the generation of weakly oxidative chlorinated species (Cl^•−/Cl₂^•−) (Equations (73) and (74)) [96]. In contrast, Wang et al. [97] have elucidated that, despite a drop in CBZ degradation kinetics in the presence of 2 mM Cl⁻, it did not significantly affect the CBZ removal rate of the Fe@C-N-S@SBA-15/PMS system

Interestingly, some studies have reported that low Cl⁻ concentrations inhibited the targeted pollutant degradation, while high Cl⁻ concentrations favored pollutant removal. Gao et al. [93] found that increasing Cl⁻ concentration from 0 to 3 mM caused a reduction in the SMX degradation rate constant from 0.10244 (min·mg·L⁻¹)⁻¹ to 0.05731 (min·mg·L⁻¹)⁻¹ by CoFe₂O₄-MCM-41 with PMS, due to the production of the less reactive species Cl⁻ and HOCl (Equations (67) and (70)). However, elevating Cl⁻ concentration to 5–10 mM significantly enhanced the SMX degradation rate constant (0.11265–0.37923 (min·mg·L⁻¹)⁻¹) by the formation of oxidative HOCl and OCl⁻ (Equations (75)–(78)). Similarly, Hou et al. [83] demonstrated that 10 mM Cl⁻ could react with PMS to produce HOCl and Cl₂ (Equations (75) and (76)) to improve NPX removal in the CoCNx@SBA-15/PMS system.

Other studies showed the negligible effects of co-existing Cl⁻ on contaminant elimination. For instance, 10 mM Cl⁻ did not affect LaCoO₃/SBA-15/PMS-mediated ATR degradation [75]. There was no significant impact on SPD degradation and SCP degradation with Cl⁻ concentration of 5–15 mM in the Cu_0.76Co_2.24O₄/SBA-15/PMS [74] and Cu_0.2Ni_0.8O/SBA-15/PMS [76], respectively.

$${\mathrm{C}\mathrm{l}}^{-}+\mathrm{S}{\mathrm{O}}_4^{•-}•\to \mathrm{C}{\mathrm{l}}^{•}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(67)

$$\mathrm{C}{\mathrm{l}}^{-}+• \mathrm{O}\mathrm{H}\to \mathrm{H}\mathrm{O}\mathrm{C}{\mathrm{l}}^{•}$$

(68)

$$\mathrm{H}\mathrm{C}\mathrm{l}\mathrm{O}{•}^{-}\to \mathrm{C}{\mathrm{l}}^{-}+• \mathrm{O}\mathrm{H}$$

(69)

$$\mathrm{H}\mathrm{C}\mathrm{l}\mathrm{O}{•}^{-}+{\mathrm{H}}^{+}\to \mathrm{C}{\mathrm{l}}^{•}+{\mathrm{H}}_2\mathrm{O} (\mathrm{A}\mathrm{c}\mathrm{i}\mathrm{d} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s})$$

(70)

$$\mathrm{C}{\mathrm{l}}^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(71)

$$\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}+{\mathrm{H}}^{+}+\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{C}{\mathrm{l}}_2+{\mathrm{H}}_2\mathrm{O}$$

(72)

$$\mathrm{C}{\mathrm{l}}^{•}+\mathrm{C}{\mathrm{l}}^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{C}{\mathrm{l}}_2^{•-}$$

(73)

$$2\mathrm{C}{\mathrm{l}}^{•-}\to 2\mathrm{C}{\mathrm{l}}^{-}+\mathrm{C}{\mathrm{l}}_2$$

(74)

$$\mathrm{C}{\mathrm{l}}^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}$$

(75)

$$2\mathrm{C}{\mathrm{l}}^{-}+\mathrm{H}\mathrm{S}{\mathrm{O}}_5^{-}+{\mathrm{H}}^{+}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{C}{\mathrm{l}}_2+{\mathrm{H}}_2\mathrm{O}$$

(76)

$$\mathrm{C}{\mathrm{l}}_2+{\mathrm{H}}_2\mathrm{O}\to \mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}+{\mathrm{H}}^{+}+\mathrm{C}{\mathrm{l}}^{-}$$

(77)

$$\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}\to \mathrm{O}\mathrm{C}{\mathrm{l}}^{-}+{\mathrm{H}}^{+}$$

(78)

As evidenced by multiple studies, the presence of HCO₃⁻ disadvantaged the degradation efficiency of pollutants [83,96,97,105]. The addition of 5 mM HCO₃⁻ reduced BPA degradation efficiency from 96.1% to 87.1% in the Co@NC-ZS/PMS system, primarily owing to the formation of less reactive HCO₃^• (Equations (79) and (80)) [95]. Similarly, Afzal et al. [75] implied that 10 mM HCO₃⁻ suppressed ATR degradation from 100% to 22% by scavenging free radicals and the generation of less reactive HCO₃^• species (Equations (79)–(81)) in the LaCoO₃/SBA-15/PMS system. Gao et al. [93] suggested that the explanation for the decreased SMX removal efficiency by adding HCO₃⁻ (0–10 mM) was not only related to the formation of less reactive HCO₃^• (Equations (79) and (80)), but also the elevated solution alkalinity induced by HCO₃⁻, which further impeded SMX degradation.

On the contrary, Huang et al. [78] demonstrated a slight enhancement in the UDMH degradation rate constant (0.2840 min⁻¹ to 0.3074 min⁻¹) by PMS with the addition of AgCo/SBA-15, because HCO₃⁻ facilitated PMS activation to generate additional SO₄^•−. He et al. [74] found that adding HCO₃⁻ (5–15 mM) accelerated SPD degradation kinetics in the Cu_0.76Co_2.24O₄/SBA-15/PMS system without altering overall efficiency, which was caused by the HCO₃⁻ oxidation (Equations (82) and (83)). Yu et al. [76] demonstrated no significant impact of 5–15 mM HCO₃⁻ on SCP removal by Cu_0.2Ni_0.8O/SBA-15 during PMS oxidation, as HCO₃⁻ minimally affected solution pH and did not react with the dominant oxidant (¹O₂).

$$\mathrm{S}{\mathrm{O}}_4^{•-}+\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{H}\mathrm{C}{{\mathrm{O}}_3}^{•}$$

(79)

$$\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+• \mathrm{O}\mathrm{H}\to \mathrm{C}{\mathrm{O}}_3^{•-}+{\mathrm{H}}_2\mathrm{O}$$

(80)

$$\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+• \mathrm{O}\mathrm{H}\to \mathrm{H}\mathrm{C}{\mathrm{O}}_3^{•}+\mathrm{O}{\mathrm{H}}^{-}$$

(81)

$$\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+\mathrm{O}{\mathrm{H}}^{-}\to \mathrm{C}{\mathrm{O}}_3^{•}+{\mathrm{H}}_2\mathrm{O}$$

(82)

$$\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}+{\mathrm{O}}_2^{•-}\to \mathrm{C}{\mathrm{O}}_3^{•}+\mathrm{H}{\mathrm{O}}_2^{-}$$

(83)

Adding NO₃⁻ generally exhibited a relatively slight inhibition on the oxidation efficiency of functionalized mesoporous silicas. A decrease was observed on the UDMH degradation rate constant from 0.2840 min⁻¹ to 0.2482 min⁻¹ as a result of the produced NO₃⁻ being less reactive (Equations (84) and (85)) [78]. The Fe@C-N-S@SBA-15/PMS system maintained approximately 95% CBZ degradation with the presence of 2 mM NO₃⁻ [97], while CBZ removal rate dropped from 100% to 77% in the Fe₃O₄@SBA-15/PDS system at a concentration of 3 mM NO₃⁻ [105]. The NO₃⁻ concentration of 5 mM slightly lowered BPA catalytic oxidation from 96.1% to 95.3% via the Co@NC-ZS-activated PMS system [95].

However, several studies showed the negligible effects of NO₃⁻ on the pollutant degradation performance, such as ATR removal by LaCoO₃/SBA-15 [75], CIP degradation by Co/N@ZS-SA [96], SMX elimination by CoFe₂O₄-MCM-41/PMS [93], SCP removal by Cu_0.2Ni_0.8O/SBA-15 [76] and SPD decomposition by Cu_0.76Co_2.24O₄/SBA-15 [74] during the catalytic PMS processes.

$$\mathrm{N}{\mathrm{O}}_3^{-}+• \mathrm{O}\mathrm{H}\to \mathrm{N}{\mathrm{O}}_3^{•}+\mathrm{O}{\mathrm{H}}^{-}$$

(84)

$$\mathrm{N}{\mathrm{O}}_3^{-}+\mathrm{S}{\mathrm{O}}_4^{•-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+\mathrm{N}{\mathrm{O}}_3^{•}$$

(85)

Research has indicated that H₂PO₄⁻ generally inhibits the degradation efficiency of the contaminant in functionalized mesoporous silica-activated persulfate systems. Wang et al. [97] revealed that introducing 2 mM H₂PO₄⁻ diminishes CBZ degradation efficiency from 100% to about 47% in the Fe@C-N-S@SBA-15/PMS process. Adding 5 mM H₂PO₄⁻ to the Co@NC-ZS/PMS system reduced BPA degradation slightly from 96.1% to 93.1%, because H₂PO₄⁻ reacted with SO₄^•− and •OH to form less reactive H₂PO₄^• (Equations (86) and (87)) [95]. However, Qin et al. [96] found that 5 mM H₂PO₄⁻ did not play an important role in CIP degradation by the Co/N@ZS-SA/PMS process, while higher concentrations (10–20 mM) of H₂PO₄⁻ enhanced CIP degradation from 82.52% to 90%, which was attributed to H₂PO₄⁻ acting as a nucleophile to accelerate PMS decomposition and ROS generation. Additionally, He et al. [74] noted a negligible effect of 5–15 mM H₂PO₄⁻ on SPD degradation during the Cu_0.76Co_2.24O₄/SBA-15/PMS process.

$$\mathrm{S}{\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{P}{\mathrm{O}}_4^{-}\to \mathrm{S}{\mathrm{O}}_4^{2-}+{\mathrm{H}}_2\mathrm{P}{\mathrm{O}}_4^{•}$$

(86)

$$• \mathrm{O}\mathrm{H}+{\mathrm{H}}_2\mathrm{P}{\mathrm{O}}_4^{-}\to \mathrm{O}{\mathrm{H}}^{-}+{\mathrm{H}}_2\mathrm{P}{\mathrm{O}}_4^{•}$$

(87)

The effects of HPO₄²⁻ varied over different persulfate systems with functionalized mesoporous silica. The introduction of 20 mM HPO₄²⁻ to the Cu-Al/MSS /PDS system reduced BPA degradation from 90% to 40%, primarily owing to competitive adsorption between HPO₄²⁻ and BPA for -OH groups on the sample [101]. Interestingly, Gao et al. [93] have reported a concentration-dependent dual effect of HPO₄²⁻ concentration on SMX removal in the CoFe₂O₄-MCM-41/PMS process. Adding 1 mM HPO₄²⁻ significantly suppressed the SMX degradation kinetics constant from 0.10244 (min·mg·L⁻¹)⁻¹ to 0.04133 (min·mg·L⁻¹)⁻¹ due to the scavenging of SO₄^•− and •OH radicals, while increasing HPO₄²⁻ concentration to 10 mM was beneficial to the SMX degradation rate constant (0.10368 (min·mg·L⁻¹)⁻¹), a result of the promotion of the generation of SO₄^•−. Additionally, negligible effects were exhibited on the SCP degradation by the Cu_0.2Ni_0.8O/SBA-15-activated PMS system when adding 5–15 mM HPO₄²⁻, due to the minimal pH alteration by HPO₄²⁻ and the non-reactivity of the dominant reactive species (¹O₂) [76]. Similarly, the presence of 10 mM HPO₄²⁻ did not influence ATR degradation significantly by the LaCoO₃/SBA-15/PMS system [75].

Co-existing SO₄²⁻ generally had adverse effects on pollutant degradation. For instance, the addition of 10 mM SO₄²⁻ prolonged the degradation time from 55 to 80 min for 100% NPX degradation by the CoCNx@SBA-15/PMS system, as SO₄²⁻ hindered PMS activation [83]. Luo et al. [95] also established a slight reduction on BPA degradation from 96.1% to 92.5% with 5 mM SO₄²⁻ in the Co@NC-ZS/PMS system. However, 10 mM SO₄²⁻ did not observably affect ATR degradation by LaCoO₃/SBA-15/PMS [75].

CO₃²⁻ also exhibited pronounced inhibitory effects on catalytic performance of functionalized mesoporous silica. Adding 10 mM CO₃²⁻ drastically reduced ATR degradation from 100% to approximately 22% in the LaCoO₃/SBA-15/PMS system, as CO₃²⁻ quenched SO₄^•− and •OH radicals to produce CO₃^•− with less oxidative potential (Equations (88) and (89)) [75]. Similarly, the addition of 3 mM CO₃²⁻ dropped CBZ degradation from 100% to 5% by PDS with Fe₃O₄@SBA-15 as a heterogeneous catalyst [105].

$$\mathrm{C}{\mathrm{O}}_3^{2-}+• \mathrm{O}\mathrm{H}\to \mathrm{C}{\mathrm{O}}_3^{•-}+\mathrm{O}{\mathrm{H}}^{-}$$

(88)

$$\mathrm{C}{\mathrm{O}}_3^{2-}+\mathrm{S}{\mathrm{O}}_4^{•-}\to \mathrm{C}{\mathrm{O}}_3^{•-}+\mathrm{S}{\mathrm{O}}_4^{2-}$$

(89)

5.7. Natural Organic Matter

Natural organic matter (NOM), a complex mixture of organic compounds formed through the physical, chemical, and biological transformation of biological residues (plants, animals, and microorganisms) in environmental matrices (soil, water, and sediments), is ubiquitously distributed in aquatic ecosystems. As the dominant NOM constituent, humic acid (HA) features a complex molecular architecture with aromatic cores and carboxyl and phenolic hydroxy groups, making it a representative proxy for studying the behavior of NOM in real aquatic environments.

The majority of studies indicated that HA exhibited an inhibitory effect on pollutant degradation in functionalized mesoporous silica-activated persulfate systems. Huang et al. [105] reported that the degradation efficiency of CBZ decreased from 100% to 27% in the Fe₃O₄@SBA-15/PDS system, as the HA concentration increased to 5 mg/L, attributed to electrophilic attacks by radicals (SO₄^•− and •OH) on electron-rich sites of NOM. Gao et al. [93] observed that the kinetic constant values of the SMX degradation rate constant from 0.02821 (min·mg·L⁻¹)⁻¹ to 0.00242 (min·mg·L⁻¹)⁻¹ by the CoFe₂O₄-MCM-41/PMS system, with HA concentration raising from 5 mg/L to 30 mg/L, resulting from the competition with SMX for ROSs and consumption of SO₄^•−. A similar phenomenon was presented for NPX degradation by CoCNx@SBA-15 [83] and CIP removal by Co/N@ZS-SA with the presence of PMS [96]. However, He et al. [74] have suggested that there was only a 5% decline in SPD degradation by the Cu_0.76Co_2.24O₄/SBA-15-activated PMS system with an HA concentration of 5–15 mM. Shao et al. [71] clarified that HA concentration plays a minor role in LVF degradation by Co@SBA-16/ceramic in the presence of PMS, because the reaction rate of HA with SO₄⁻· was much lower than that with •OH.

6. Degradation Pathways and Toxicity

Although functionalized mesoporous silica materials/persulfate systems exhibit high degradation efficiency towards target pollutants in water, their mineralization capacity is limited, resulting in the generation of numerous oxidative intermediates and by-products. These oxidative intermediates and by-products could be detected via liquid chromatography-mass spectrometry or gas chromatography-mass spectrometry. The differing attack sites of ROS on pollutant molecules lead to the formation of distinct intermediates and by-products in different degradation systems. SMX in the CoFe₂O₄-MCM-41/PMS system was degraded via the S-N bond cleavage, hydroxylation, S-C bond cleavage, and isoxazole ring opening to achieve mineralization, generating thirteen potential intermediates and by-products (Figure S1a) [93]. The S-N bond was broken by electrophilic species (SO₄^•−, •OH, O₂^•− and ¹O₂). Hydroxylation was induced by •OH attack, and both S-C bond and isoxazole ring were broken by SO₄^•−. The ROS generated in the Cu_0.76Co_2.24O₄/SBA-15/PMS system attacked the benzene ring, C-N and N-C-N bond of SPD to produce eight identified intermediates, with these intermediates were further oxidized to CO₂ and H₂O. Notably, no SPD peaks were detected in mass spectrometry, proving nearly complete degradation (Figure S1b) [74]. SCP was degraded by a break of electron deficient chemical bonds, C-S bond, C-N bond, and an additional reaction in the Cu_0.2Ni_0.8O/SBA-15-activated PMS system. Subsequently, ring-opening and cleavage occurred, finally forming small molecules (Figure S1c) [76]. Fluoroquinolones are degraded through distinct routes. Shao et al. [71] identified four main routes for LVF elimination in the Co@SBA-16/ceramic/PMS system, including LVF dehydrogenation, hydroxylation, piperazine epoxidation and ring opening, and quinolone ring opening, from there detecting 11 intermediates/by-products (Figure S1d). The CIP degradation pathways were different in the CFMCM-41/PMS (Figure S1e) [106] and Co/N@ZS-SA/PMS (Figure S1f) [96] systems. The former removed CIP by a stepwise oxidation of piperazine ring and decarboxylation, while the latter eliminated CIP through the piperazine ring opening, oxidation of tertiary amine in the piperazinyl, and defluorination. The Cu-SBA-15/PMS mineralized TC via dimethylamino cleavage, amide group loss/hydroxy dehydration, and six-membered ring opening (Figure S1g) [60]. One of the degradation pathways of TC in the Fe-C/N-SBA-15/PMS system was dissociation of the dimethylamino group. The other method was demethylation of the dimethylamino group, removal of acylamino group, and dehydration from hydroxyl group. These intermediate products then opened a six-membered ring to form lower molecular weight products (Figure S1h) [61]. Hou et al. [83] suggested that the degradation pathways of NPX are primarily categorized into hydroxylation and decarboxylation in the CoCNx@SBA-15/PMS system, followed by gradual breakdown into CO₂ and H₂O through processes such as dimerization and ring-opening mineralization (Figure S1i). Wang et al. [97] have revealed that CBZ degradation in the Fe@C-N-S@SBA-15/PMS system was achieved via three pathways: (i) •OH attack on electron-rich N, resulting in a loss of an amide group; (ii) an electron transfer to produce a C-centered radical cation; and (iii) hydroxylation (Figure S1j). Eighty products were detected in the Co@MCM-41/PMS/AZRS system, and there were six degradation pathways in the AZRS removal, including hydroxylation, de-sulfonation, ring-opening, de-sulfonation/breaking, de-sulfonation/hydroxylation and de-sulfonation/ring-opening/hydroxylation (Figure S1k) [91]. MO was degraded in the FeCo-MCM-41/PMS system via three routes: desulfonation, benzene ring-amino bond cleavage, and azo bond cleavage (Figure S1l) [63]. Endocrine disruptors like phenol degraded in Co/SBA-15/PMS via sequential oxidation and ring-opening. The formed oxalic acid was hard to oxidize and might be the main source of residual TOC [70]. CuO/MSS/PDS (Figure S1m) [100] and Co@NC-ZS/PMS (Figure S1n) [95] degraded BPA via four (dehydrogenation, hydroxylation, C-C cleavage, and β-fragmentation) and two (radical oxidation and dihydroxylation) routes, respectively. Both generated intermediates were finally mineralized to CO₂ and H₂O. Huang et al. [78] found that UDMH degradation was achieved via primary amine H-substitution and tertiary amine nitrogen attack routes in the AgCo/SBA-15/PMS system. It formed 10 intermediates and gradually mineralized to CO₂, H₂O, and N₂ (Figure S1o). DFT calculations also showed that tertiary/primary amine nitrogen were most susceptible to electrophilic ROS attacks; therefore, UDMH degradation started with these sites.

Degradation intermediates and by-products generated during degradation processes may pose risks to ecological security, therefore it is urgent to assess their environmental risk. Eco-toxicity evaluation generally involves both theoretical simulation and practical experiment. The Toxicity Estimation Software Tool (TEST) 5.1 is employed to simulate the bioaccumulation factor, developmental toxicity, and mutagenicity of the target pollutant and its intermediates/by-products. Gao et al. [93] found that the predicted bioaccumulation factors for all of the intermediates/by-products were lower than that of SMX in the CoFe₂O₄-MCM-41/PMS system. The developmental toxicity of most intermediates/by-products was also predicted to be lower. With the exception of one product, all other intermediates/by-products were predicted to be non-mutagenic (mutagenicity value < 0.5). The Ecological Structure Activity Relationships (ECOSAR) 2.2 program is applied to predict the acute and chronic toxicity of the target pollutant and its degradation intermediates/by-products towards fish, daphnia, and green algae. Shao et al. [71] reported that certain individual intermediates/by-products exhibited acute toxicity to daphnia and green algae, as well as chronic toxicity to fish, daphnia, and green algae during the LVF degradation by the Co@SBA-16/ceramic/PMS system. However, it is important to note that these intermediate products were ultimately transformed into non-toxic substances as the reaction proceeded. A similar phenomenon was observed in the degradation of SMX by the CoFe₂O₄-MCM-41/PMS system [93].

7. Reusability and Stability

This section summarizes the different regeneration methods of functionalized mesoporous silica materials and their reusability in activating persulfate for pollutant degradation. The stability of functionalized mesoporous silica materials was further analyzed through leaching experiments and characterization techniques, as presented in

Table 3. Reusability of functionalized mesoporous silica materials.

Catalyst	Persulfate	Pollutant	Regeneration Method		Reusability	Leaching Concentration	Ref.
Cu0.76Co2.24O4/SBA-15	PMS	SPD	Solvent method	Deionized water	1st run: 99.20%, 5th run: 98.79%		[74]
NG/NH2-MCM-41	PMS	BPA		Deionized water	5th run > 85%		[102]
Mn-Co-MCM-41	PMS	RhB		Deionized water	6th run > 90%		[64]
Co/SBA-15-Cl	PMS	Phenol		Deionized water	2nd run: 86.6%		[58]
AgCo/SBA-15	PMS	UDMH		Deionized water	1st run: 0.2860 min−1, 2nd run: 0.2155 min− 1, 3rd run: 0.1889 min− 1, 4th run: 0.1776 min− 1, 5th run: 0.1732 min− 1		[78]
Co@NC-ZS	PMS	BPA		Deionized water	5th cycle: 84.4%		[95]
GO-MCM-Fe	PDS	LVF		Ethanol	1st run: 97.28%, 2nd run: 96.87%, 3rd run: 94.90%, 4th run: 89.40%, 5th run: 85.13%	Fe leaching: 0.99%	[120]
Co/N@ZS-SA	PMS	CIP		Ethanol and deionized water	3rd run: no significant loss	Co leaching: ~0.55 mg/L per cycle	[96]
Sr2FeO4/SBA-15	PMS	SPD		Ethanol and deionized water	5th cycle > 94.0%	Sr leaching < 3.5 mg/L, Fe leaching < 3.0 mg/L	[77]
Cu0.2Ni0.8O/SBA-15	PMS	SCP		Ethanol	1st run: 99.27%, 4th run: 88.06%	Cu leaching: 0.197 μg/L, Ni leaching: 0.153 μg/L	[76]
CF/SBA-15	PMS	RhB		Ethanol and deionized water	1st run: 95%, 4th run: 63%	Co leaching < 30 μg/L	[72]
Co@SBA16/ceramic	PMS	LVF		Ethanol and deionized water	1st run: 78%, 3rd run: 28.6%		[71]
Cu-BMS	PDS	MB		Isopropyl alcohol and deionized water	1st cycle: 93.5%, 4th cycle: 84.3%, 8th cycle: 75.7%, 12th cycle: 70.4%	Cu leaching: 0.105 mg/L (1st), 0.088 mg/L (4th), 0.026 mg/L (8th), 0.017 mg/L (12th)	[112]
Fe3O4@SBA-15	PDS	CBZ		NaOH solution and deionized water	1st run: ~100%, 6th run: 89%	Fe leaching: 0.78 mg/L per cycle	[105]
g-C3N4/MCM-41	PMS	AO7		KOH solution and deionized water	1st run: 96.8%, 2nd run: 93%, 3rd run: 85%		[94]
CoMg/SBA-15	PMS	RhB	Pyrolysis	Calcination (400 °C)	25th cycle > 94%	1st cycle: Co leaching: 80 μg/L, Mg leaching: 4.3 mg/L. 5th cycle: Co leaching: 25 μg/L, Mg leaching < 100 μg/L.	[79]
CF/SBA-15	PMS	RhB		Calcination (500 °C, 3 h)	10th run > 84%	Co leaching < 72.1 μg/L, Fe leaching < 35 μg/L.	[72]
MnOx/SBA-15	PMS	BPB		Calcination (823 K, 2 h)	6th cycle: 98.4%.	Mn leaching < 80 μg/L per cycle.	[81]
CFMCM-41	PMS	CIP		Calcination	1st run: k = 0.174 min−1, 2nd run: k = 0.045 min−1, 3rd run: k = 0.041 min−1	Co leaching = 31 μg/L, Fe leaching = 15.5 μg/L.	[106]
CFS	PMS	Acid orange II	Solvent and pyrolysis method	1–5 cycles: Ethanol and deionized water, 6th cycle: calcination (400 °C)	1st run: >90%, 2nd run: 80%, 3rd run: >40%, 5th cycle: ~10%, 6th cycle: >80%		[117]
CoFe2O4-MCM-41	PMS	SMX		1–5 recycles: deionized water, 6th recycle: calcination (550 °C, 5 h)	1st cycle: 96.82%, 5th recycle: 81.57%, 6th recycle: 95.87%	Co leaching = 0.037–0.061 mg/L Fe leaching = 0.057–0.091 mg/L	[93]
CoCNx@SBA-15	PMS	NPX		1–5 cycles: deionized water, 6th cycle: Calcination (400 °C, 4 h, N2)	1st cycle: 100% in 20 min, 2nd cycle: 100% in 80 min, 3rd cycle: 100% in 80 min, 4th cycle: 89.7% in 80 min, 5th cycle: 75.1% in 80 min, 6th cycle: 100% in 80 min.		[83]
g-C3N4/MCM-41	PMS	AO7		1–3 cycles: deionized water, 4th cycle: calcination (300 °C, 1 h)	1st cycle: 96.8% in 30 min, 2nd cycle: 93% in 120 min, 3rd cycle: 85% in 180 min, 4th cycle: 77% in 120 min		[94]

.

7.1. Reusability

The reusability of materials is a crucial criterion for evaluating catalyst performance. The multiple cycling test is an effective approach to assess the reusability of catalysts. For instance, a significant reduction (86.6%) in the phenol degradation rate constant was observed in the second run of the Co/SBA-15-Cl/PMS system. This deactivation was primarily ascribed to cobalt leaching, which originated from the location of Co₃O₄ particles on the external surface of the SBA-15 [58]. The degradation efficiency of BPA by CuO/MSS-activated PDS decreased from 92% to over 80% with the Cu leaching concentration of 0.073 mg/L after five cycles. The decline in degradation efficiency was attributed to the reduction of catalytically active sites caused by Cu leaching and the deposition of Cu(OH)₂ [100]. Huang et al. [78] indicated that a decline in the rate constant of UDMH degradation from 0.286 min⁻¹ to 0.1732 min⁻¹ in the AgCo/SBA-15/PMS system can be ascribed to the leaching of metal species, blocking of active sites by intermediates, and unavoidable loss of catalyst. The Fe@C-N-S@SBA-15/PMS system achieved a 100% degradation efficiency for CBZ in the 1st run but dropped to less than 55% in the 4th cycle. The decline primarily resulted from the loss of ferrous species and accumulation of CBZ and its produced intermediates deposited on the catalyst surface [97]. The relatively high Fe leaching concentration from Fe@C-N-S@SBA-15 (2.34 mg/L) suggests that further attention should be paid to the catalyst deactivation mechanism to facilitate the development of catalysts with more abundant and stable reactive interfaces. A decrease in the apparent degradation rate of CIP from 0.174 min⁻¹ to 0.041 min⁻¹ after the 3rd run by the CFMCM-41/PMS system can be attributed to pore blockage and loss of adsorption sites within the catalyst structure by CIP and its intermediates [106]. Similar phenomena were found in the MO and UDMH removal achieved by the FeCo-MCM-41/PMS [63] and AgCo/SBA-15/PMS systems [78], respectively.

Spent functionalized mesoporous silica catalysts are typically collected, treated, and reused for pollutant degradation during cycling experiments. Regeneration strategies for these functionalized mesoporous silicas are conventionally classified into solvent method and thermal treatment. Solvent-based regeneration methods are further subdivided into deionized water, organic solvent, and inorganic alkaline solution washing, among which deionized water rinsing represents a predominantly employed regeneration approach. The SPD degradation efficiency only decreased from 99.2% (1st cycle) to 98.79% (5th cycle) by Cu_0.76Co_2.24O₄/SBA-15 in the presence of PMS [74]. The NG/NH₂-MCM-41/PMS system retained over 85% efficiency over five operational cycles for BPA degradation [102]. A reduction from 98% to 90% for RhB removal after six successive runs was observed in the Mn-Co-MCM-41-activated PMS oxidation system [64]. In contrast, the Co/SBA-15-Cl /PMS system exhibited a significant drop in the phenol degradation rate constant from 0.1725 ppm/min in the 1st cycle to 0.0232 ppm/min in the 2nd cycle, attributed to the weakened interaction between Co and the SBA-15 matrix [58]. For AgCo/SBA-15 washed with deionized water, the degradation rate constant by PMS for UDMH decreased from 0.2860 min⁻¹ (1st use) to 0.1732 min⁻¹ (5th use), mainly due to the leaching of metal species, blockage of active sites by reaction intermediates, and material loss during the recovery and washing process [78]. The BPA degradation efficiency decreased from 95.0% to 84.4% after five cycles by PMS using Co@NC-ZS as a heterogeneous catalyst, attributed to minor catalyst loss and metal ion leaching [95].

The organic solvent also serves as a common regeneration method, primarily because deionized water is insufficient to remove contaminants adsorbed onto the catalyst surface. Ethanol is one of the most frequently employed organic solvents. For an ethanol-washed GO-MCM-Fe activating PDS system, the degradation efficiency of LVF decreased from 97.28% to 85.13% after five regeneration cycles [120]. Co/N@ZS-SA/PMS, washed with deionized water and ethanol alternately, showed no significant activity loss over three cycles [96]. Similarly, the combined washing with deionized water and ethanol was implemented to remove adsorbed organic matter and residual reactants from the Sr₂FeO₄/SBA-15 surface. Its PMS activation system maintained SPD degradation efficiency above 94% even after the 5th cycle, compared with 99% in the initial use [77]. Regeneration with anhydrous ethanol led to a decline in the SCP degradation efficiency for the Cu_0.2Ni_0.8O/SBA-15/PMS system, decreasing from 99.27% to 88.06% after four cycles. This was primarily due to surface passivation by electron-deficient functional groups derived from intermediates, which impeded electron transfer from the catalyst to PMS [76]. Similarly, sequential washing with ethanol and distilled water resulted in a reduction of RhB degradation efficiency (from 95% to 63% after four cycles) for a PTFE membrane-supported CF/SBA-15-activated PMS system. This was caused by the progressive accumulation and deposition of carbonaceous intermediates blocking the active sites of the catalyst [72]. Furthermore, regeneration via alternating deionized water and ethanol washes induced a pronounced drop (49.4%) in the Co@SBA-16/ceramic/PMS system for LVF degradation after only three cycles. This rapid performance loss was mechanistically attributed to inherently weak interfacial interactions between Co species and the SBA-16 framework, progressive aggregation of Co₃O₄ nanoparticles during reaction cycles, and diminished accessible active sites. Notably, a performance assessment occurred in a continuous-flow fixed-bed reactor configuration. While this operational mode inherently induced the loss of metal active sites due to fluidic shear forces, it more accurately simulated hydraulic scouring conditions encountered in real water treatment applications [71]. Cu-BMS regenerated through deionized water and isopropyl alcohol washing demonstrated a decline in PDS activation efficiency for MB degradation, decreasing from 93.5% to 70.4% after twelve consecutive reaction–regeneration cycles [112].

Alkaline solution washing is another significant regeneration method for spent catalysts. Fe₃O₄@SBA-15 regenerated by NaOH solution exhibited a decline in CBZ degradation efficiency by PDS from nearly 100% to 89% after six cycles, primarily due to the transition of Fe(II) species in Fe₃O₄ to Fe(III) [105]. For g-C₃N₄/MCM-41 regenerated via deionized water washing, the PMS activation efficiency towards AO7 degradation declined from 96.8% within 30 min during the initial cycle to 85% after 180 min following three reuse cycles. Subsequent alkaline treatment using KOH restored the degradation efficiency to 93%. This performance recovery was mechanistically attributed to the depletion of N-C-O functional groups—identified as the primary cause of activity decay—with hydroxide treatment effectively reestablishing their optimal surface concentration [94].

Thermal treatment is able to effectively eliminate intermediates/pollutants adhered to catalyst surfaces that solvent washing could not remove, restoring persulfate activation performance to near-original levels. The calcination at 400 °C could remove carbonaceous deposits occupying catalytic active sites on the CoMg/SBA-15 and enhance covalent bonding between Co/Mg and SBA-15. Thus, the CoMg/SBA-15/PMS system can maintain an RhB degradation efficiency of over 94% after 25 cycles [79]. Similarly, CF/SBA-15, calcined at 500 °C for 3 h, achieved more than 84% RhB degradation after 10 consecutive runs with the presence of PMS [72]. Yang et al. [81] demonstrated that MnO_x/SBA-15, regenerated through 2 h calcination at 823 K, maintained a high efficiency of BPB degradation, retaining 98.4% degradation efficiency within PMS in the 6th cycle. This performance recovery was mechanistically linked to the pyrolysis-driven removal of oxygen-containing adsorbates from spent catalysts.

Due to the high efficiency in regenerating catalysts of pyrolysis, some studies combined solvent washing with thermal regeneration. Specifically, solvent washing was employed during the initial regeneration cycles, and then transitioned to pyrolysis once a marked decline in contaminant degradation efficiency was observed. The degradation efficiency of acid orange II via PMS activation by CFS declined markedly from more than 90% to about 10% after five cycles, following regeneration with deionized water and ethanol washing, due to the adsorption of intermediate organic species onto the active sites of the catalyst. Remarkably, calcination at 400 °C in air prior to the 6th cycle restored the PMS activation efficiency for acid orange II degradation to a level comparable to that of the fresh catalyst [117]. Gao et al. [93] demonstrated that CoFe₂O₄-MCM-41 regenerated solely with deionized water exhibited a diminishing efficiency from 96.82% to 81.57% after the 5th regeneration by PMS for SMX degradation. Due to the extremely low leaching of Fe and Co ions, the performance loss was ascribed to the deposition of reaction intermediates on the catalyst surface, which obstructed active sites and hindered ROS generation. Subsequent calcination at 550 °C for 5 h removed accumulated intermediates and restored SMX degradation efficiency to 95.87%, a performance similar to that of the fresh catalyst. CoCNx@SBA-15/PMS achieved a complete NPX removal in 20 min initially, but a decline was shown in subsequent cycles (100%, 100%, 89.7%, and 75.1% in cycles 2–5). Subsequently, CoCNx@SBA-15 pyrolyzated at 400 °C under a N₂ atmosphere activated PMS to reach a 100% NPX removal in 80 min, effectively removing the accumulated acidic oxygen-containing functional groups from the material surface. However, the performance remained inferior to that of the fresh sample, potentially owing to the irreversible loss of N functional groups in CoCNx@SBA-15 [83]. In contrast, the AO7 degradation efficiency via PMS activation by g-C₃N₄/MCM-41 regenerated with deionized water washing decreased from 96.8% within 30 min to 85% in 180 min after the 3rd cycle. Subsequent thermal treatment at 300 °C for 1 h slightly improved the efficiency to 77% in 120 min, which was slightly higher than the result obtained in the 3rd cycle [94].

7.2. Stability

Metal leaching is a key indicator for assessing catalyst stability. Loading metal oxides on mesoporous silica effectively mitigates metal leaching, thereby preserving the activity of the catalyst over multiple cycles. Afzal et al. [75] found the Co leaching concentrations of LaCoO₃/SBA-15 were 0.04, 0.02, and 0.01 mg/L during the 1st, 2nd, and 3rd catalytic cycles, respectively, confirming the structural stability. Hu et al. [72] observed that the Co and Fe leaching concentrations remained below 72.1 μg/L and 35 μg/L, respectively, over ten consecutive regeneration cycles of CF/SBA-15. He et al. [74] documented leaching concentrations of 0.350 μg/L for Cu(II) and 0.364 μg/L for Co(II) in Cu_0.76Co_2.24O₄/SBA-15. The Cu leaching concentrations of Cu-BMS were only 0.105, 0.088, 0.026, and 0.017 mg/L in the 1st, 4th, 8th, and 12th cycles, respectively, indicating a decreasing trend over successive uses. The Co/SBA-15 catalyst exhibited sustained Co leaching below 85 μg/L throughout 25 regeneration cycles [70], while the Cu leaching remained approximately constant at about 0.1 mg/L per cycle for Cu-Al/MSS [101].

Furthermore, characterization of the materials before and after reaction provides critical insights into their structural stability. Morphological integrity was evidenced by SEM/TEM in Cu-BMS (no significant agglomeration or fragmentation) [112], Co/SBA-15 (despite surface particles, no significant shedding) [70], Co@SBA-16/ceramic (an increase of particle size due to agglomeration) [71], CFS (spinel particles intact despite minor silica shell damage) [117], LaCoO₃/SBA-15 (uniform morphology without loss) [75], Cu-Al/MSS (no significant change in the peak) [101], CoMg/SBA-15 (retained agglomerated morphology) [79], and Mn-Co-MCM-41 (preserved cubic morphology) [64]. XRD patterns consistently preserved crystalline structures, as demonstrated by unchanged characteristic peaks in Cu-BMS [112], Co/SBA-15 [70], CoFe₂O₄-MCM-41 [93], γ-Fe₂O₃@2Co-MCM-41 [65], CFS (partial spinel structure) [117], CoMg/SBA-15 (Co₃O₄/MgO phases) [79], Cu-Al/MSS (CuO/Al₂O₃ phases) [101], and g-C₃N₄/MCM-41 [94]. FT-IR analyses further indicated unaltered framework bonds in g-C₃N₄/MCM-41 [94], FeCo-MCM-41 (Si-O-Si/M-O bonds) [63], and Mn-Co-MCM-41 [64]. XPS patterns of Co@SBA-16/ceramic showed consistent surface oxygen composition and Co satellite peaks, despite Co₃O₄ particle growth [71]. Fe-C/N-SBA-15 exhibited unchanged XPS peaks and binding energy post-reaction [61], while Mn-Co-MCM-41 [64] and CuO/MSS [100] exhibited stable porosity via N₂ adsorption–desorption. Additional evidence included high metal retention (>96% for γ-Fe₂O₃@2Co-MCM-41 by ICP [65]) and preserved surface functionality (retained basic sites in CoMg/SBA-15 by CO₂-TPD [79]), collectively affirming structural integrity post-regeneration.

8. Practical Application Potential

Degradation performance in real water matrices was investigated to evaluate the practical application potential of functionalized mesoporous silica materials in environmental remediation, providing a theoretical basis for the optimization and engineering application of environmental remediation materials. Current studies have typically assessed the catalytic performance of functionalized mesoporous silica by conducting experiments using water collected from real environmental matrices (e.g., tap water, river water, lake water, wastewater treatment plant effluent), and employing experimental conditions identical to those used in deionized water systems. Notably, target contaminants were artificially elevated to mg/L concentrations, which substantially exceed environmentally relevant levels observed in natural waters (typically ng/L to μg/L). Afzal et al. [75] demonstrated that the LaCoO₃/SBA-15/PMS system achieved comparable ATR removal (>98% within 6 min for 10 mg/L ATR) in drinking water, tap water, river water, and seawater, indicating negligible inhibition from co-existing ions and NOM. Luo et al. [95] have clarified that the degradation efficiency of BPA (C₀ = 20 mg/L) in tap water by the Co@NC-ZS/PMS system was similar to that in deionized water. However, the degradation rate decreased for sea water potentially due to active site competition by seawater constituents and anion interference. Despite this, the BPA degradation rate in tap water and sea water by Co@NC-ZS reached over 90% within 30 min. A significant decline was observed in the CFMCM-41/PMS system for CIP removal in real water samples such that the degradation dynamics constants of CIP (C₀ = 5 mg/L) were 0.174 min⁻¹, 0.109 min⁻¹, 0.069 min⁻¹, and 0.003 min⁻¹ in deionized water, tap water, river water, and secondary effluent, respectively [106]. This was mainly due to the ROS scavenging by the Cl⁻ added during disinfection in tap water and the competition for ROS by NOM and anions in river water and secondary effluent. Similarly, Huang et al. [105] found lower degradation efficiencies of CBZ (C0 = 10 mg/L) in mineral water and tap water than deionized water by Fe₃O₄@SBA-15/PDS system, a resulting of the quenching effect caused by matrix components. Gao et al. [93] revealed a negative relationship between the SMX removal efficiency (C₀ = 10 mg/L) by the CoFe₂O₄-MCM-41-activated PMS system and total organic carbon (TOC) concentration presented in the aqueous matrix. These results establish competitive interference from NOM as a dominant factor contributing to reduced catalytic efficiency. To rigorously evaluate the practical applicability of CoFe₂O₄-MCM-41 for SMX remediation, river water exhibiting the highest TOC content was selected as the test matrix and the initial SMX concentrations were chosen at environmentally relevant levels (50, 80, and 100 μg/L). The results indicate the comparable degradation performance for SMX in river water and deionized water under identical conditions, achieving a complete degradation within 1 min. This phenomenon can be attributed to adequate ROS production in the CoFe₂O₄-MCM-41/PMS system, effectively overcoming matrix complexity. This outcome substantiated the practical application potential of the system for SMX remediation in real water bodies.

9. Conclusions and Prospects

In this review, we systematically summarized and categorized literature on functionalized mesoporous silica materials for the activation of persulfate to remove pollutants from water environment. Active components can be loaded onto mesoporous silica via co-condensation, impregnation, grinding–calcination, hydrothermal synthesis, or chemical precipitation methods. The dispersion state and microenvironment of active components on mesoporous silica can be controlled by adjusting synthesis parameters, thereby influencing the catalytic ability. Metallic-functionalized mesoporous silica primarily activated persulfate through metal redox cycles (valence changes of metal active sites), while nonmetallic-functionalized materials relied on nonmetal groups. The metallic–nonmetallic co-functionalized mesoporous silica utilized the above pathways to drive persulfate. Therefore, targeted pollutants were degraded via radical pathways (SO₄^•−, •OH and O₂^•−), non-radical pathways (¹O₂, electron transfer), or both. Additionally, confinement effects further contributed to pollutant degradation. The catalytic performance of catalysts was influenced by reaction conditions and external factors. Increasing catalyst and oxidant dosages appropriately improved degradation efficiency, whereas higher initial pollutant concentrations reduced the pollutant removal. The effects of temperature, initial solution pH, coexisting anions (e.g., Cl⁻, HCO₃⁻, NO₃⁻, H₂PO₄⁻, HPO₄²⁻, SO₄²⁻, CO₃²⁻), and HA varied over different functionalized mesoporous silica/persulfate systems. Functionalized mesoporous silica materials exhibited excellent stability and could be regenerated through solvent and thermal treatments, or a combination of them. Moreover, they exhibited excellent adaptability to pollutants in complex water environments.

Despite the significant potential in the field of functionalized mesoporous silica materials in wastewater treatment, several critical challenges remain for future development. Currently, the functionalized mesoporous silica catalyst-driven persulfate activation focuses on the regulation of the materials to obtain maximal catalytic efficiency as well as excellent stability and reusability. Although costs could be reduced to some extent by decreasing the dosage and multiple uses of materials in this way, no literature has reported on the preparation and usage costs of functionalized mesoporous silica materials. Current synthesis methods primarily rely on laboratory-scale chemical reagent approaches, including high-purity silica sources (e.g., TEOS), organic templates (e.g., CTAB and P123), and active species precursors (e.g., metal nitrates, metal sulfates, or metal complexes). Although these methods are capable of controlling meso-structures, they suffer from problems of high raw material costs and energy consumption, toxicity of template agents, and complex procedures for template removals. To promote practical applications, there is an urgent need to develop green extraction routes based on low-cost silica sources derived from agricultural waste (e.g., rice husks, diatomite) and industrial byproducts (e.g., fly ash, metallurgical silica slag), thereby reducing economic and environmental loads through resource recycling strategy. Additionally, regulatory principles on governing coordination environments of active sites anchored mesoporous silica materials remain inadequately understood. The location (internal pores or external surfaces), existing forms (metal oxides, single-atom metals, or metal-nonmetal groups), and loading amounts directly influence mass transfer efficiency and catalytic activity toward pollutants. However, existing studies on the effects of synthesis conditions, such as template type, silica source impurities, and reaction parameters (e.g., pH, temperature, calcination atmosphere, and temperature) on the coordination environment of active sites mostly remain at the level of empirical description. The lack of in situ tracking of dynamic evolution processes make it difficult to regulate active sites precisely by adjusting synthetic parameters. Furthermore, the missing structure–activity relationship between the chemical microenvironment of active sites and pollutant degradation mechanisms results in a gap in the knowledge of correlations associated with the “synthesis conditions—active site structure—pollutant degradation efficiency.” This gap impedes the rational design of functionalized mesoporous silica materials tailored to specific needs. Future research should undertake in situ characterization techniques (e.g., X-ray absorption fine structure, Raman spectroscopy) with multiscale theoretical calculations (e.g., density functional theory simulations, mechanistic learning) to elucidate the dynamic evolution mechanisms of active sites under different synthetic conditions and their regulatory effects on reaction pathways to lower costs and toxicity. Furthermore, coupling catalysts with low-energy input sources (e.g., visible light, ultrasound) could further promote efficient radical generation, enabling lower catalyst requirements.

Catalyst recovery technology still faces significant bottlenecks. Only a small fraction of functionalized mesoporous silica materials containing magnetic elements could be separated by magnetic field, while most materials still rely on traditional centrifugation/filtration techniques, which suffer from severe material loss, high energy consumption, and secondary pollution risks. To address this issue, future efforts should not only be made on developing magnetic functionalized mesoporous silica materials, but also on the immobilization technology of nano-functionalized mesoporous silica materials, such as those loaded onto hydrophilic membranes or gels to integrate functional reactor and upscale systems (e.g., fluid and fixed bed reactors). These approaches could significantly enhance recovery efficiency, prolong service life, and reduce the operational costs of materials.

There remains a significant gap in terms of pollutant removal between the experiment conditions in the laboratory and actual application environments. This discrepancy manifests in the following key aspects. Firstly, current studies have generally adopted single-pollutant systems prepared with deionized water or spiked water samples with high concentrations of target antibiotics (mg/L), which showed a difference of orders of magnitude in comparison with pollutant concentrations in real water bodies (ng/L-μg/L). This distorts assessments of reaction kinetics and active site occupancy. Secondly, the real aquatic water matrices are inherently complex systems governed by the synergistic effects of multiple factors, such as temperature, pH, coexisting ions, NOM, and organic pollutants. However, this complexity is often oversimplified in laboratory studies, typically reduced to single-factor investigations or neglected altogether. Consequently, this simplification hinders the accurate understanding of the correlations between multiple influencing factors and the removal efficiency of target pollutants. The application of functionalized mesoporous silica-activated persulfate systems for the removal of pollutants in actual water samples from various regions has not been sufficiently explored. Comprehensive studies are still needed to systematically assess the degradation efficiency and practical applicability of these systems under diverse water quality conditions characteristic of different geographical areas. Most importantly, most degradation experiments employ batch modes, lacking continuous operation performance evaluation of catalysts, failing to reflect hydraulic conditions and long-term stability in actual water treatment processes. The differences between batch experiments and real water bodies lead to severe distortion in catalyst performance evaluation. In addition, although current studies reported that metal leaching concentrations of most functionalized mesoporous silica were at trace levels (ng/L to μg/L), the long-term accumulation of these leached metals—even at such low concentrations—may still pose significant risks to environmental safety. To address this challenge, validation methods should be developed that more accurately simulate real aquatic conditions. The continuous-flow systems could be applied to replicate realistic hydraulic loading rates and dynamic water quality variations. Coupling advancing in-situ monitoring and feedback control technologies could be used to detect trace-level pollutants and dynamic adjustment of conditions in response to changes in wastewater composition. It is essential to assess the catalytic performance of the systems across various real water matrices from different geographical regions, in order to determine its broader applicability and robustness under environmentally relevant conditions. Key strategies include the adoption of experimental design methodologies such as the response surface methodology (RSM) or Taguchi orthogonal arrays, which efficiently account for interactive effects between operating parameters (e.g., catalyst loading, oxidant dose, pH, temperature and reaction time) while optimizing reaction conditions and lowering costs and toxicity. It is also imperative to place greater emphasis on monitoring metal leaching concentrations under prolonged operational conditions and to further investigate the long-term toxicological effects associated with such exposure. Hence, the practical catalytic activity of catalysts, persistent challenges like metal leaching, surface contamination, and catalyst corrosion under sustained oxidative stress and operational economic cost could be accurately evaluated through these systematic experiments. This could provide reliable foundations for engineering applications and promote the sustainable management of contaminated water bodies.