Zeolite Modification for Optimizing Fenton Reaction in Methylene Blue Dye Degradation

Abstract

Textile dyes often prove resistant to conventional wastewater treatment processes because of their complex molecular structures. Advanced oxidation methods, such as the Fenton reaction, have thus been recognized as a promising approach for environmental remediation by decomposing these pollutants. This work aimed to study the efficacy of modified zeolites as catalysts in the Fenton reaction for dye degradation, with a particular emphasis on techniques for modifying zeolites and incorporating iron. The zeolite ZSM-5 was selected as the parent structure and underwent desilication and acid treatment procedures. Iron was introduced into the zeolite structure via two distinct methods: ion exchange and mechanochemistry. The modified zeolites with incorporated iron were evaluated in terms of their crystallinity, textural properties, and iron content before being used to degrade methylene blue solutions through the Fenton reaction. The reaction was monitored using UV-Vis spectroscopy, while the experimental outcomes were analyzed using pseudo-first-order and pseudo-second-order kinetic models. The research findings indicate that different treatment methods led to varying impacts on the zeolite properties, which in turn influenced the kinetic results. Moreover, it was observed that an enhancement in the degradation process can be achieved through the harmonious balance between a high iron content, increased mesoporosity (to facilitate diffusion), and adequate crystallinity (essential for maintaining structural integrity).

1. Introduction

The textile industry is one of the largest contributors to environmental pollution globally, with dyeing and finishing processes generating large volumes of wastewater that contain high concentrations of dyes. The chemical structure of dyes makes them difficult to remove through traditional methods, such as coagulation, flocculation, and sedimentation. Many dyes are synthetic compounds with complex molecular structures, which can bind to particles in the water and resist degradation. This means that even after treatment, these pollutants can still be present in the environment, posing a risk to aquatic life and human health. While conventional methods have limitations, advanced oxidation processes, like the Fenton reaction, offer promising solutions for reducing the environmental impact of the textile industry [1].

The Fenton reaction, named after the British chemist Henry Fenton [2], uses hydrogen peroxide (H₂O₂) and a catalyst, typically iron(II) or iron(III), to generate highly reactive non-selective hydroxyl radicals (^●OH). These radicals can degrade a wide range of organic pollutants, including dyes [3]. The reactions are outlined in Equations (1) and (2):

Fe²⁺ + H₂O₂ → Fe³⁺ + OH⁻ + ^●OH

(1)

Fe³⁺ + H₂O₂ → ^●OOH + Fe²⁺ + H⁺

(2)

The ^●OH initiates reactions with organic substances, transforming them into organic radicals. These radicals then undergo a sequence of oxidation reactions, ultimately producing secondary and tertiary metabolites until mineralization (Equation (3)).

RH + ^●OH → H₂O + ^●R → Degradation → Mineralization

(3)

The use of zeolites in the Fenton reaction for the degradation of dyes has emerged as a promising approach in environmental remediation. Zeolites are a class of microporous, aluminosilicate minerals, which are commonly used in a variety of industrial applications, including water treatment, gas separation, and, especially, catalysis. They are also renowned for their exceptional ion exchange capacities, a property that makes them ideal supports for metals such as iron, copper, and manganese, essential components in the Fenton reaction. Due to their ordered structure and porosity, zeolites can act as effective supports for metal species, promoting high dispersion inside the porous structure, which enhances the catalytic efficiency when compared to solely metallic catalysts, which have lower catalytic activity, stability, and recyclability [4].

Several key factors can change the effectiveness of zeolites in the Fenton reaction. One major factor is the pore size and shape of the zeolite framework; this influences the adsorption capacity, crucial for concentrating pollutant molecules within the porous structure, maximizing the contact between pollutant molecules and the reactive hydroxyl radicals generated during the Fenton process [5,6]. Additionally sufficient large pores allow for the efficient mass transfer of reactants and intermediates within the zeolite framework [7].

Further research is needed to optimize the use of zeolites in this application and to explore the potential of different types of zeolites as catalyst supports. Therefore, in this work, we propose studying different aspects of zeolite properties and their impact in the efficiency of the Fenton reaction. Our approach involves investigating a commercial ZSM-5 zeolite (Figure 1), which has previously been used as an individual support medium for the Fenton reaction [8]. However, no relevant studies have explored its modification and the subsequent implications for reaction dynamics.

A portion of the initial sample underwent a desilication treatment and was further subjected to an additional acid treatment. As extensively reported in the literature, desilication is one of the most common procedures performed to modify the porosity of zeolites, leading to an enlargement in microporosity and the creation of mesoporosity through the selective removal of Si from the zeolite framework [10]. In many cases, the debris resulting from desilication are deposited as extra-framework species and could block the access to the inner porosity; thus a subsequent acid treatment is often required [11]. This strategy has previously been employed by the authors in different studies [12,13]. The objective of this work is to produce materials with different textural properties and study how these parameters can affect the metal ion dispersion and the access of the reactive species during the methylene blue (MB) degradation reaction.

To enhance dispersibility and reduce ion leaching, previous studies have incorporated metals, such as Fe, Cu, and Mn, into zeolite synthesis via hydrothermal co-precipitation [14] and the sol-gel method [15]. However, in our work, the metal incorporation is carried out after the zeolite post-treatments using two distinct techniques: the ion exchange method and the mechanochemistry procedure.

MB belongs to the phenothiazine family and was chosen as a representative model molecule for studying the degradation behavior by the heterogeneous Fenton reaction. This molecular structure (Figure 2) allows us to study the fundamental mechanisms of oxidative degradation, which can then be applied to more complex dye structures.

Additionally, it has a characteristic absorption spectrum, making it easy to detect and track its concentration during the degradation process. All these attributes make the MB molecule a common choice in Fenton reaction studies [16,17,18].

2. Materials and Methods

2.1. Materials and Reagents

The zeolite used as starting material was ZSM-5 (CBV 3024E, Lot 2200-99), with a SiO₂/Al₂O₃ molar ratio = 30, acquired from Zeolyst International (Kansas, KS, USA) in ammonium form. To obtain the protonated form, a thermal treatment (calcination) was performed in a muffle furnace (Nabertherm, B170, Lillenthal, Germany). In the first 3 h, the temperature increased from room temperature to 550 °C at a rate of 5 °C/min, and for the remaining 3 h, the temperature was held constant at 550 °C.

Methylene blue (MB) was supplied by Merck Chemicals (Darmstadt, Germany), the 30% hydrogen peroxide solution was supplied by Carlo Erba (Emmendingen, Germany), and the sodium thiosulphate pentahydrate 99% was supplied by Panreac (Barcelona, Spain). The chemicals used for the zeolite treatments were acquired from Merck Chemicals (Darmstadt, Germany) and were used as received.

2.2. Zeolite Modifications

The desilication of the zeolite ZSM-5 was performed with NaOH according to the procedure described in [19], with a concentration of 0.4 M. The zeolite was stirred in an NaOH solution using 30 mL of liquid per gram of solid on a heating plate with stirring and temperature control (IKA C-MaHS7, Staufen, Germany) at 60 °C for 30 min. The solid was recovered through vacuum filtration, washed, and dried in an oven at 110 °C. Subsequently, an ion exchange was performed to replace some cations (Na⁺) that may have been exchanged during the previous treatment to ensure that the zeolite returned to their full protonated form. The exchanges were carried out at 80 °C for 6 h using a 2 M NH₄NO₃ in a mixture containing 25 mL of liquid per gram of solid. The solid was collected afterwards under the same conditions as previously mentioned and then calcined at 500 °C for 3 h in a muffle furnace. Finally, to remove some extra-framework Al species that may cause the partial occlusion of the pore openings, an acid treatment was performed by suspending the solid in 0.1 M HCl solution at 70 °C for 3 h in a mixture containing 30 mL of liquid per gram of solid. The solid was recovered as previously described. The modified samples were designated accordingly with the suffix “DAT”, referring to the desilication (D) + acid treatment (AT) (see Figure 3).

2.3. Metal Introduction

The metal was introduced by two methods: classic ion exchange (IE) and mechanochemistry (M). Both procedures involved preliminary calculations aimed at incorporating iron(III) ions (Fe³⁺) into a mass percentage roughly equivalent to 0.7 wt%, a typical and average value used in previous studies by the authors [20].

In the case of the ion exchange, the zeolite samples were suspended in a 3.65 mM solution of Fe(NO₃)₃·9H₂O (>99% Merck) using a ratio of 35 mL of solution per g of zeolite at room temperature for 24 h under agitation using a stirring plate (P-Selecta, Agimatic-S, Barcelona, Spain). Then, the reaction medium was collected and centrifuged at 6000 rpm for 5 min (Hermle, Z206A, Gosheim, Germany). The supernatant was separated from the solid and the latter was placed in the oven at 80 °C for 24 h.

In the mechanochemical method, the zeolite was mixed in a beaker with the appropriate amount of Fe(NO₃)₃•9H₂O. The mixture was then put in a steel vessel of the ball mill (VWR Star Beater, Avantor, Radnor, PA, USA) with 1/3 of its capacity occupied; 5 steel metal spheres with a diameter of 3 mm were used to perform the mechanical action. The frequency used was 10 Hz and the treatment time was 10 min. The reaction conditions in the ball mill, the number of balls, diameter, frequency, and test time were previously optimized. All the modified samples were calcined at 350 °C for 4 h (with a temperature ramp of 5 °C/min) in the muffle. Figure 3 represents the sample identification according to the type of metal loading methodology.

2.4. Catalyst Characterization

The samples were characterized by various techniques: powder X-ray diffraction; N₂ adsorption isotherms at −196 °C; and chemical analysis to quantify the iron content. The crystallinity percentage was estimated using the data obtained from the X-ray diffraction (Philips Analytical PW 350/60 X’Pert PRO with X’Celerated detector, Almelo, The Netherlands). This involved collecting diffractograms at room temperature under CuKα radiation through continuous scanning between 5 and 40° (2θ) with a step size of 0.017° (2θ) and a 40 s dwell time per step. The textural properties of the samples were studied using nitrogen adsorption isotherms measured at −196 °C with ASAP 20 automated equipment (Micromeritics Instruments Corporation, Norcross, GA, USA). Prior to the analysis, the samples were degassed in a vacuum greater than 10⁻² Pa for 3 h at 300 °C.

The iron content of the samples was determined by ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry). The analyses were performed at “Laboratório de Análises”, IST, Lisboa, Portugal.

2.5. Catalytic Experiments

The following procedure was applied in each of the seven batch experiments: 10 mg of each of the solid samples was accurately weighed and placed in 50 mL stoppered flasks. Then, 25 mL of the dye solutions, that were previously prepared at a concentration of 7 ppm, were added to the flasks and then closed. The selected concentration assured that after reaching adsorption equilibrium the dye concentration allows the reaction to be monitored for a period of at least one hour before the dye is completely degraded. The flasks were then immersed in a thermostatic bath at 40 °C (Julabo MP, Seelbach, Germany), which was placed on a multiposition magnetic stirrer (Selecta Multimatic 9-S, Barcelona, Spain), and adsorption was allowed to come to equilibrium for at least 1 h. This equilibrium step was fundamental because our objective was to isolate the catalytic effect from the zeolite dye adsorption phenomenon, which is always present. This equilibrium stage was previously verified to occur within a time frame of less than 1 h; only after achieving this we initiated the reaction.

When formerly starting the reaction, the solution’s pH was adjusted to a value of around 5 using a HCl 0.5 M solution; this pH value was proposed by previous authors as optimal for dye degradation [21]. The reaction was started by adding 0.5 mL of a 70 mM H₂O₂ solution to each of the stoppered flasks. At a pre-established time, in each one of the stoppered flasks, a small quantity (aprox. 0.05 g) of solid sodium thiosulphate was added to stop the reaction.

Before the analysis, the catalyst was removed from the dye solution by centrifugation (Hermle, Z206A, Gosheim, Germany), using a speed of 6000 rpm for 10 min. Aliquots of the supernatant solution were taken and their absorbances were measured in a double-beam spectrophotometer (Jasco V530, Tokyo, Japan) using deionized water as a reference and a cell of either 10 or 2 mm optical length. Prior to the assays, a calibration curve was performed at 663 nm using dye solutions with the concentration range needed to obtain absorbances that were between 0.15 and 1.0, to obey the Beer–Lambert law. Each data point resulted from an average of at least three individual aliquots/scans, assuring a standard deviation below 5%.

3. Results and Discussion

3.1. Zeolite Characterization

Figure 4 depicts the XRD patterns for the different catalysts, including the parent and modified samples of ZSM-5.

As expected, all the samples reproduced the XRD patterns typical of the parent zeolite. To account for some loss of crystallinity as a consequence of the treatments, the degree of crystallinity (C_XRD) was estimated following the procedures reported in the ASTM D5758-01 method [22]. The resulting crystallinity percentage was calculated by dividing the cumulative area of selected peaks of the different samples by the area of the respective parent, ZSM-5, using Peak-fit 4.12 software (Grafiti Inc., Palo Alto, CA, USA). The results are depicted in

Table 1. Crystallinity, textural properties, and iron content of the zeolite samples.

Sample	CXRD 1 (%)	Vsuper (cm3 g−1)	Vultra (cm3 g−1)	Vmicro 2 (cm3 g−1)	Vmeso 3 (cm3 g−1)	Aext (m2 g−1)	Fe (wt.%)
ZSM-5	100	0.05	0.10	0.15	0.07	41	-
FeZSM-5_IE	100	0.04	0.10	0.14	0.07	43	0.29
FeZSM-5_M	100	0.04	0.10	0.14	0.07	41	0.61
ZSM-5_DAT	100	0.05	0.10	0.15	0.14	88	-
FeZSM-5_DAT_IE	94	0.05	0.10	0.15	0.13	36	0.33
FeZSM-5_DAT_M	36	0.03	0.10	0.13	0.12	62	0.65

¹ Degree of crystallinity calculated from powder X-ray diffraction patterns, using parent ZSM-5 as reference. ² Total, V_micro, ultra, V_ultra, and super, V_super, micropore volumes and external surface area, A_ext, were quantified through the application of α_s method; ³ Mesoporous volume, V_meso = V_total − V_micro, where the total volume (V_total) corresponds to the amount of N₂ adsorbed at p/p^o ≈ 0.95.

. As can be observed, the parent zeolite loaded with Fe by the ion exchange (IE) and mechanochemistry (M) methods, i.e., FeZSM-5_IE and FeZSM-5_M, respectively, preserved the crystallinity. Concerning the modified samples through desilication + acid treatment (DAT), the parent sample, ZSM-5_DAT, retained the initial crystallinity; however, the method used for Fe-loading had different effects. In the case of the FeZSM-5_DAT_IE sample, there was practically no crystallinity loss, whereas for the FeZSM-5_DAT_M sample, a significant loss was verified, since this sample was previously fragilized by the alkaline + acid procedures before entering the ball mill. This led to structural disruption within the zeolite framework upon subjecting it to the mechanical forces of the milling process.

The N₂ adsorption–desorption isotherms at −196 °C for the different samples within the relative pressure range of 0–1 are depicted in Figure 5. It must be mentioned that, to allow for an ease analysis of the isotherms’ configuration, the data of samples FeZSM-5_IE and FeZSM-5_DAT_IE are not represented since the curves are coincident with those of samples FeZSM-5_M and FeZSM-5_DAT_M, respectively. For all samples, the curves can be classified as type I + IV isotherms [23]. The hysteresis at moderately elevated pressures (P/P₀ = 0.5−0.95) provided evidence of some mesoporosity, probably generated from the aggregation of small crystals in the case of ZSM-5, which became more important for DAT samples because of the alkaline + acid treatment. The α_s method was used to quantify the textural parameters, taking as reference the isotherm obtained from a non-porous silica [24]. The total micropore volume (V_micro) was determined using the α_s method by extrapolating the linear portion of the α_s graph, defined by experimental points corresponding to α_s > 1. The mesopore volume (V_meso) was obtained through the difference between V_total and V_micro, where V_total is the amount of nitrogen adsorbed at a relative pressure of 0.95, based on Gurvitch’s rule [24]. Using α_s method, it was possible to distinguish between the volume of narrow micropores, V_ultra, corresponding to pores with dimensions less than 0.7 nm, and the volume of supermicropores, V_super, which corresponds to pores between 0.7 and 2 nm. The value of V_ultra was determined by extrapolating the line defined by the experimental points obtained at low relative pressures above 0.02 to the origin. By subtracting V_ultra from the total micropore volume (V_micro), the value of V_super was obtained.

The textural properties presented in Table 1 show that the metal loading in the parent ZSM-5 had no impact on the textural parameters, no matter the loading method used. For the alkaline + acid modified samples, the foreseen mesoporosity development is demonstrated by the values of V_meso, which, in the case of ZSM-5_DAT, became double that of ZSM-5. Contrary to what was observed in the case of ZSM-5, the method used to introduce Fe into the ZSM-5_DAT sample had different repercussions on the textural parameters. While the ion exchange resulted only in a small decrease in the V_meso and the A_ext, the mechanochemical treatment affected both the larger micropores and the mesopores. In this case, the decrease in V_meso was associated with a smaller decrease in A_ext, which may indicate the formation of smaller particles, created by the milling of a fragilized structure, in agreement with the previously reported crystallinity loss.

To complement the analysis of the N₂ adsorption data, the mesopore size distributions obtained by Hybrid Density Functional Theory (DFT Plus ^® V2.01 ASAP 2010 V5.00), considering pores with cylindrical shape, are also displayed in Figure 5 (bottom). The curves are, in all the cases, almost parallel, revealing that the Fe-introduction procedures had no effect in the mesopore size distribution.

Table 1 also depicts the iron content of the different samples. This parameter is a factor that directly influences catalytic behavior, since iron is introduced as an active site for reactions, and its content affects both activity and adsorption.

The results show that the targeted amount of iron (0.7 wt.%) was achieved when the metal ion was introduced through mechanochemistry but was far from the desired Fe loading when the ion exchange method was used. In this case, the high Si/Al ratio of the zeolite (15) likely contributes to its lower aluminum content, which reduced the availability of sites for ion exchange reaction. The same behavior was replicated in the case of the DAT samples, evidencing the benefits of ion metal loading through mechanochemistry.

3.2. Heterogeneous Fenton Reaction

To evaluate the contributions of the zeolite samples, comparative experiments were conducted while keeping key experimental parameters, such as the catalyst mass and the hydrogen peroxide (H₂O₂) concentration, constant. The degradation of MB over the parent zeolite, ZSM-5, and in the presence of only H₂O₂ were carried out as control tests. In the sole presence of H₂O₂, a negligible MB conversion was observed, as the reaction is effective only when Fe ions are anchored at the zeolitic supports.

The kinetic curves for the degradation of MB are depicted in Figure 6. Despite observable differences in catalytic behavior, as depicted by the kinetic curves, quantifying their relative efficiency requires converting these curves into numerical data. This involves deriving kinetic constants from the observed kinetics.

The kinetics were investigated using pseudo-first-order and pseudo-second-order models, since despite the high number of steps, Fenton reactions are frequently simplified to those two models. The kinetic constants were obtained from nonlinear regressions using Table Curve 2D 5.0 software (Grafiti Inc., Palo Alto, CA, USA).

The pseudo-first-order rate constant assumes that the rate of the reaction depends on the concentration of a single reagent. This model is presented in Equation (4).

$${\displaystyle \frac{\left[c\right]}{\left[C_0\right]}}=e^{-k_{ap1}t}$$

(4)

where [c] is the dye concentration, [C₀] is the initial dye concentration, k_ap1 is the pseudo-first-order kinetic constant, and t corresponds to time.

The pseudo-second-order rate constant is often related to adsorption processes, where the reaction rate depends on both the reagent and the surface sites [25]. The form of this model is presented in Equation (5).

$${\displaystyle \frac{\left[c\right]}{\left[C_0\right]}}={\displaystyle \frac{1}{k_{ap2}{[c}_0]t+1}}$$

(5)

where k_ap2 is the pseudo-second-order kinetic constant, and all the other symbols have the same meaning as described before.

The numerical results of the pseudo-first-order and pseudo-second-order constants are depicted in

Table 2. Pseudo-first-order and pseudo-second-order model application to MB degradation kinetic curves and associated statistical descriptors.

Samples	Pseudo-First-Order Model
	kap1 (min−1)	R2	sFit	F	N
FeZSM-5_IE	0.038 ± 0.002	0.984	0.043	366	7
FeZSM-5_M	0.120 ± 0.020	0.909	0.099	60	7
FeZSM-5_DAT_IE	0.061 ± 0.008	0.970	0.068	97	4
FeZSM-5_DAT_M	0.037 ± 0.002	0.990	0.032	529	6
Samples	Pseudo-Second-Order Model
	kap2 (ppm−1 min−1)	R2	sFit	F	N
FeZSM-5_IE	0.019 ± 0.003	0.940	0.0830	93	7
FeZSM-5_M	0.052 ± 0.006	0.978	0.0480	276	7
FeZSM-5_DAT_IE	0.100 ± 0.002	0.999	0.0007	8971	4
FeZSM-5_DAT_M	0.023 ± 0.003	0.975	0.0530	198	6

, along with the statistical parameters for the nonlinear regressions to evaluate the quality of adjustments, namely, the standard error of the constants, the coefficient of determination (R²), the fit standard error of the regression (s_Fit), the Fisher–Snedecor parameter (F), and the number of data points used (N).

Analyzing the kinetic results, the pseudo-first-order constants are higher than the pseudo-second-order constants. It is also possible to verify that the best model equation in the description of the kinetics varies from sample to sample.

According to some authors, the successful application of the pseudo-first-order kinetic model in Fenton’s reaction occurs because kinetics modeling considers only the initial phase of degradation, instead of the entire process [26]. Our results support this hypothesis: the reactions that better fitted the pseudo-first-order model were the ones that did not reach a plateau, FeZSM-5_IE and FeZSM-5_DAT_M.

The pseudo-second-order model usually reflects a closer description of the kinetic process due to its ability to capture the interaction between reactive species (like hydroxyl radicals) and the zeolite (iron-loaded active sites) [27]. This is also true in our case, even if we tried to “isolate” the adsorption process by allowing the dye adsorption equilibrium to be reached before starting the reaction. The relative magnitude of the pseudo-first-order and pseudo-second-order constants, and their occasional inversion, can be influenced by factors like the initial dye and H₂O₂ concentration, the specific structure of the zeolite, the iron content, and the experimental conditions used. Studies have shown that varying these factors, such as in heterogeneous Fenton or photo-Fenton processes, can lead to differences in the reaction kinetics, especially with complex dye molecules [28,29].

To better understand these results, it is also fundamental to correlate the kinetic results with zeolite properties; so, considering the results of Table 1 and Table 2, it is possible to see that the samples with high crystallinity, such as FeZSM-5_M, and FeZSM-5_DAT_IE, show higher rate constants (k) in both the pseudo-first-order and pseudo-second-order kinetics. FeZSM-5_DAT_M, with only 36% crystallinity, shows lower constants. The FeZSM-5_IE sample, despite the high crystallinity, shows lower constants. This suggests that while lower crystallinity may diminish catalytic efficiency, other effects (discussed below) may partially compensate. High crystallinity is associated with well-defined structures and with consistent and ordered channel networks, which probably improves the diffusion of reactants (such as hydrogen peroxide and dye) towards the sites with iron. Lower crystallinity, as observed in some of the modified samples, may indicate structural degradation or partial amorphization, potentially disrupting active sites or pore structures and decreasing the catalytic efficiency in the Fenton reaction.

In relation to the remaining textural parameters, there is not enough variability in V_ultra, V_super, and V_micro to justify the differences. This variability is only differentiated for V_meso between the modified and non-modified samples. In the Fenton reaction, mesoporosity helps to increase the diffusion of larger organic molecules and the accessibility of these reactants to iron sites, and this effect can be seen in the catalysts FeZSM-5_IE and FeZSM-5_DAT_IE, where the increase in mesoporosity may be one of the causes for the increase in the rate constant. The apparent contradiction when analyzing FeZSM-5-M and FeZSM-5_DAT_M is justified by the fact that the variation of the rate constant is multifactorial and, in this case, there is a parallel strong decrease in the crystallinity between these samples.

A higher iron content would be expected to increase the rate of the Fenton reaction by increasing the number of catalytic sites. However, it is not evident what amount of iron makes its increase catalytically irrelevant. Additionally, it cannot be excluded that there may be some phenomena that counteract this effect, such as an excess of iron that can lead to the formation of large iron aggregates, decreasing the dispersion of active Fe species, for example, in the mechanochemical process. This aggregation can decrease the accessibility of active sites by obstructing the pore openings; thus, a high iron content can increase non-productive side reactions. An overview of the kinetics results show that the iron content appears to have a positive correlation with kinetic rate constants when we compare FeZSM-5_IE and FeZSM-5_M. In the case of FeZSM-5_DAT_M, despite its low crystallinity, this catalyst still maintains a relatively high kinetic constant that can be attributed to its high iron content.

3.3. Regeneration Assays

The ability to reuse the catalysts was investigated by selecting two samples, FeZSM-5_IE and FeZSM-5_M, after a first catalytic run. These catalysts were submitted to two consecutive catalytic runs under the same reaction conditions upon regeneration at 400 °C for 4 h (heating rate 5 °C/min), in a muffle, after each run. The results on Figure 7A represent the MB degradation percentage after 1 h. The iron percentage present in each sample and each cycle was also quantified and is presented in Figure 7B.

As can be observed, the catalysts prepared through the IE or M methods show a high percentage of degradation (85–95%), indicating a good efficiency of the Fenton reaction during two consecutive cycles. This is because the amount of iron available is sufficient to effectively catalyze the breakdown of H₂O₂ into ^•OH radicals, which rapidly attack the dye molecules. For the third cycle, the percentage of degradation decreased significantly, especially in the case of the FeZSM-5_M sample (70% for FeZSM-5_IE and 46% for FeZSM-5_M). The reduction in the amount of iron (Fe%) after each cycle may indicate leaching (loss of iron into the solution). This leaching effect involves a mechanism including iron ion exchange, diffusion, and dissolution steps.

The decrease in the concentration of available iron limits the amount of hydroxyl radicals that can be generated, leading to the lower degradation efficiency observed in the third cycle. However, the iron reduction is not decisive in the second cycle, meaning that in this cycle, there are enough iron sites to promote the reaction. This hypothesis was validated to ensure that the leached iron metals in the solution did not contribute significantly to the MB degradation. Control experiments were carried out using Fe³⁺ solutions with a concentration equivalent to a zeolite iron content of 0.7 wt.%; the results showed significantly slower reactions with homogeneous catalysis, with kinetic constant values 100 times lower than those of heterogeneous catalysis.

The slight increase in the degradation efficiency of FeZSM-5_IE from the first to the second cycle (from 85% to 94%) could be indicative of more exposed iron active sites, probably from the washing and removal of debris from the pore openings [19].

4. Conclusions

In this work, two methods were used to incorporate Fe (III) into commercial and modified (desilication + acid treated) ZSM-5 zeolite. The results obtained show that mechanochemistry is a simple and effective method with which to load the desirable amount of metal ions, without a significant impact on the structure and texture of the samples. On the other hand, the samples modified through desilication + acid treatment developed the desired mesoporosity. Nevertheless, in this case the metal ion-loading method must be the classic ion exchange since the incorporation through mechanochemistry leads to a severe loss of crystallinity, accompanied by a decrease in the textural parameters. The properties of the samples were then connected to the kinetic constants that were obtained through the modeling of the kinetic data from the dye degradation reaction against both pseudo-first-order reactions and pseudo-second-order reactions. The data obtained showed that the kinetics of this process were influenced by multiple factors. The higher degree of crystallinity positively affected the catalytic activity, as well as the mesoporous volume, enhancing reactant accessibility, which is pivotal for larger molecules or reactions that are more sensitive to diffusion. An increase in the iron content may have led to a rise in the number of active sites, with a positive impact on the rate of dye degradation. However, there appears to be a threshold beyond which its impact diminishes based on regeneration assays. The tendency of iron to leach from the zeolite structure was noted, but the reactions sustained a high rate of degradation during two consecutive cycles.