Photocatalytic Degradation of Azo Dyes over Magnetite-Based Catalyst: Kinetic, Thermodynamic and Structure-Activity Studies

1. Introduction

Due to environmental concerns, the global demand for lesser consumption of water and lowers discharge of industrial and domestic wastewater is a major challenge to environmental sustainability [1,2]. An important example is the consumption of excessive amounts of water and synthetic dyes by textile industrial processes, which constitutes harmful effects on human health and ecosystems, resulting in a huge amount of non-biodegradable organic compounds in the environment [2,3,4]. In this respect, hazardous wastewater treatment processes for removing organic pollutants are required to overcome their excess and improve the discharged water quality.

Efficient processes for rapid removal of organic contaminants, such the use of advanced oxidation processes (AOPs), have become ubiquitous in remediation methods for clean-up of wastewater by organic contaminants degradation to mitigate their adverse consequences on the environment [5,6]. Moreover, the state-of-the-art in AOPs process shows the chemical oxidation of dyes through in situ generation of highly reactive oxygen species e.g., hydroxyl radicals (^●OH), superoxide radicals (O₂^●−) and singlet oxygen (¹O₂) by mechanistic routes of photocatalysis, ozonation, electrocatalysis, oxidation by Fenton reaction, and Fenton-like processes [7,8,9,10,11]. One of the promising route for dyes degradation is the Fenton photocatalytic process, which has outstanding advantages for textile wastewater removal with rapid reaction rate and time [7,12,13]. Especially, the use of iron based catalysts in Fenton-like processes gives photogenerated electrons to rise the Fe³⁺/Fe²⁺ conversion rate, while light is used for decolorization of dyes contaminants [13,14].

One of the difficulties associated with the utilization of the benchmark photo-Fenton catalysts such as α-Fe₂O₃, Fe₃O₄ and FeOOH among other iron-based oxides, is a high photogenerated electron-hole pair complexation rate, being considered to be a crucial factor that influences the low accessibility of the active sites and the light absorption efficiency of the catalytic performance in these processes [15,16,17,18]. It is acknowledged that the current state of Fe²⁺ ions present in magnetite is not yet sufficient to reliably assign H₂O₂ gives the aforesaid hydroxyl and hydroperoxyl radical (HO₂^●) and/or superoxide radical anions to allow the oxidation of Fe²⁺ to Fe³⁺ along with the consumption of ^●OH radicals by the proper Fe²⁺ ion addressing high efficiencies to oxidize the organic pollutants [19].

On the other hand, it should be noted than use of magnetite as catalysts allow to reduce the unit operation and accelerate the recovery of the magnetic catalyst which makes wastewater treatment more inexpensive. The improvement of magnetite properties seems to be very promising for obtaining a highly stable catalyst during reactions for oxidative degradation of dyes at circumneutral pH [15,16,17,18].

Therefore, it is considered important to understand the magnetite properties to optimize the catalyst activity in heterogeneous photo-Fenton-like process for the oxidation of azo dyes, however, the effect of functional groups present in the dye molecule on the heterogeneous phase oxidation has been little explored in the literature. In the present study, the degradation kinetic and thermodynamic, the H₂O₂ and pollutant concentrations, pH, the catalyst amount and the relationship between the structures of the dyes and their photocatalytic characteristics were examined for the degradation de azo compounds e.g., Acid Red 18 (AR18), Acid Red 66 (AR66), and Orange 2 (OR2) on magnetite.

2. Results and Discussion

2.1. Heterogeneous Photo-Fenton-Like Reaction

Figure 1 shows the AR 18 azo dye photodegradation performances over the Fe₃O₄ catalyst. The effects of catalyst dosage in Fenton-like reaction on the degradation efficiency of AR 18 azo dye compared with other reaction systems are investigated.

Preliminary catalytic runs to analyze the catalyst dosage variations on the efficient degradation of AR 18 over magnetite catalyst are shown in Figure 1a. It may be noted that the AR18 degradation slight increase from 58 to 88% from catalyst dosage of 0.5 to 1.4 g.L^-1, when using an initial AR 18 of ca. 60 mg.L^-1 within 120 min, but that of 2.0 g.L^-1 is slightly decreases achieving 83%. The enhancement in degradation efficiency may be attributed to the combined effects of direct oxidation by UV light (photolytic mechanism) and the contribution of this radiation to the formation of ^●OH radicals via decomposition of H₂O₂ in an aqueous medium (photochemical mechanism) and/or on the surface of the catalyst (photocatalytic mechanism), according to equations 1 to 3. On the other hand, the slight reduction in catalytic activity for a dosage of 2.0 g.L^-1 may be associated with difficulty in penetrating light into the reaction medium due to the amount of catalyst in suspension. Nevertheless, such a rapid drop in the percentage degradation is not reflected in any alteration of the magnetite structure.

Blank runs conducted with 60 mg.L^-1 of AR 18 at pH 3.0 in 120 min using magnetite as catalyst reveal that the use of H₂O₂ and UV radiation alone results a degradation of the azo dye of less than 10%. Due to the absence of the catalyst, H₂O₂ and UV irradiation offer limitations to the reaction occurrence, where the AR18degradation is meaningless. Even though the photochemical process using UV/H₂O₂ without a catalyst may affect AR18 degradation with ~ 80% for 60 mg.L^-1 and ~ 45% for 200 mg.L^-1 (Figure 1b), the activity is comparatively modest for a Fenton-like process in the presence of magnetite with maximum degradation of 95%, under the conditions used here.

Based on the abovementioned results, the catalyst dosage is set at 1.4 g.L^-1 in the presence of H₂O₂/UV, which is suitable to undertake a systematic study of the AR18 azo dye photodegradation over the Fe₃O₄ as a catalyst. The effects of various Fenton-like reaction parameters on the AR 18 azo dye degradation efficiency are evaluated (Figure 2).

The influence of pH for Fenton and Fenton-like reaction as a parameter to elucidate the availability of Fe²⁺ to decompose H₂O₂ and generate the hydroxyl radicals in Fenton and Fenton-like reaction over iron-based catalysts with optimum pH in the range of 2.5-4.0 is evaluated [20,21,22].

The degradation efficiencies at the different pH tested are related to the zeta potential on the magnetite surface, which were + 8.4 mV at pH = 3.0, -6.26 mV at pH = 6.0 and -21.7 mV at pH = 8.0. The AR18 solution has an initial pH of 6 and under this condition the degradation efficiency is only 7.5% in 30 min of reaction, while a conversion of 32% can be achieved at pH = 3.0 for the same reaction time. After 60 min, the degradation efficiency improves and activities around 20% are only observed at natural pH (6.0) and slightly alkaline pH (8.0). On the contrary, at pH = 3.0 the degradation efficiency increases up to 60.3%. In 120 min, degradation efficiencies were 88% in acidic medium and approximately 39% and 38% at pHs = 6.0 and 8.0, respectively (Figure 2a). At pH 8.0 the magnetite catalyst is observed to be stable and there is no significant change in the degradation efficiency compared to that observed at almost neutral pH (pH = 6.0). In general, at pH values close to 3.5, azo dye molecules are effectively degraded by ^●OH radicals through the decomposition of H₂O₂ on the catalytic surface of magnetite [11]. Furthermore, the oxidation potential of the ^●OH /H₂O redox couple is greater at acidic pH values, which favors the degradation efficiency. At higher pH values (alkaline medium), the oxidizing effect decreases due to the reactions between ^●OH radicals and H₂O₂) [11,23]. Additionally, in an alkaline medium, the photocatalytic treatment of Acid Red 18 on WO₃ microspheres decorated with goethite showed a 16% reduction in degradation when the pH was increased from 1.89 to 9.64. This behavior was attributed to electronic repulsion between the anionic dye molecule and the ^-OH groups attached to the catalyst surface, as can also be related in this study [24].

Because of the initial H₂O₂ concertation greatly affects the AR18 degradation under acidic conditions [22], to understand the photocatalytic trend, a wide range of H₂O₂ concentrations should be considered. Figure 2b illustrates that the degradation of the azo dye increased linearly from 42.1 to 94.1% with H₂O₂ concentration increment in the range of 10-60 mg.L^-1 within 120 min. With further initial H₂O₂ up to 100 mg.L^-1, the degradation efficiency diminished between 2.2% e 9.4%. This might seem counterintuitive, because larger amounts of H₂O₂ would likely generate more ^•OH radicals to degrade the azo dye, However, with an excess of H₂O₂, the rate of ^•OH recombination increases (^•OH + ^•OH → H₂O₂) and the parallel reaction of ^•OH consumption by H₂O₂ is favored to form HO₂^• or O₂^- (H₂O₂ + ^•OH → H₂O + HO₂^•), which are less reactive species than ^•OH [25,26].

In addition, it is commonly accepted that the effect of azo dye concentration on Fenton and Fenton-like oxidations is limited at higher concentrations of AR18 due to the effective hydroxyl radicals formation on the magnetite catalyst [11,22,27]. Accordingly, Figure 2c shows a quite impressive almost complete AR18 degradation efficiency for a concentration below 80 mg.L^-1. As expected, the azo dye concentration between 150 and 200 mg.L^-1 gives only 76.1% and 51.5% suggesting that Fe²⁺ species of magnetite can still be present, but its availability might be low, which may be related to the adsorption on the catalyst surface of several intermediate products formed during the oxidative process, including aromatic compounds, naphthalenes and their hydroxylated and nitrified derivatives, naphtodiols and phthalic acid derivatives [28]. Hence, with the chosen magnetite catalyst loading, 200 mg.L^-1 was used for further kinetic and thermodynamics studies of AR18, AR16 and OR2.

The requirement for a stable catalyst in the Fenton-like reaction can be understood based on the recycling experiments. Figure 2d indicates that there are six cycles of consecutive uses of magnetite without any loss of degradation efficiency. At the end of the reaction, a steady state plateau about 90-94% is reached for the initial AR18 concentration of 60 mg.L^-1, at pH of 3.0 in 120 min, with no Fe species leaching as determined by AAS chemical analyses. These results are comparable to those in literature reports on spinel isostructural ferrite-based catalysts [11,29,30].

2.2. Kinetic and Thermodynamic of Degradation

According to the findings, reaction kinetics plays an important role in determining the rate of Fenton photocatalytic degradation processes [11,29].

The pseudo-first-order model was applied for the AR18, AR66 and OR2 azo des with similar structures. Figure 3 depicts the effect of reaction temperature in the range of 25–60 ◦C. At 25^oC,all azo dyes undergo higher degradation efficiency in shorter reaction time owing to elevated dye concentrations, evidencing a major collision rate between azo dyes and ^•OH radicals. With increasing of reaction temperature from 40 to 60 ◦C, an increment in reaction rate is observable for all azo dyes, at short reaction times. As the reaction proceeds, the degradation efficiency caused by the decreased collision rate between azo dyes intermediate compounds and hydroxyl radicals. In agreement, increased temperatures favor major collisions between the dye and OH radicals. In agreement, increased temperatures favor higher collisions between dye and OH radicals, although the rate is not affected with respect to the change in temperature within the range of longer reaction times [13,31,32]. The degradation of azo dyes follows the order: OR2 > AR66 > AR18, with the azo dye OR2 reacting with better efficiency at the three temperatures investigated.

The Arrhenius plots for the photo-Fenton degradation of azo dyes are depicted in Figure 4. The k₁ (min^-1) and E_a (kJ.mol^-1) parameters are given in Table 1. The Arrhenius plots of the natural logarithm of k₁ is linearly correlated with the reaction temperature, as shown in Figure 4.

Photodegradation of the azo dyes on the magnetite surface is relied upon to improve the degradation efficiency. The examined kinetic data show a high correlation coefficient of 0.97 with a good fitting between calculated and experimental values suggesting that the Fenton-like process follows first order kinetics. In addition, the reaction rates of color removal and degradation efficiencies increase with increasing temperature, with the Fenton oxidation reaction being suitable. In fact, the activation energy of AR18 is 36.2 kJ.mol^-1, comparable with the degradation efficiency of Fe-based catalysts on Acid Red 18 azo dye in other studies [21,29,31,32,33,34]. For OR2, the calculated activation energy is 25.3 kJ.mol^-1, comparable with literature reports on magnetite-based catalysts [35,36,37,38]. These values indicate that the photocatalytic Fenton reaction rate can be controlled by the intrinsic kinetics at the surface of the magnetite catalyst.

The parameter determination of Gibbs (ΔG), entropy (ΔS) and enthalpy (ΔH) for the AR18, AR66 and OR2 photodegradation, based on the van't Hoff equation in Figure 5. The obtained data are summarized in Table 2.

The entalphy (ΔH) values obtained for the AR18, AR66 and OR2 azo dye are all similar to those of [39] , in the case of Acid Red 27. The ΔS values for AR18, AR66 and OR2 are 0.196, 0.157 and 0.158 kJ.mol^-1.K^-1, respectively, suggesting that the reaction is more spontaneous with increasing temperature [40]. The ΔG values for the OR2 degradation in the range of temperatures investigated are higher than those found for AR66 and AR18, which corroborates with the degradation order and the reaction kinetics (OR2 > AR66 > AR18).

This behavior can be explained in terms of molecular effects related to the polar group natures, possible steric hindrance and, to some extent, to the type of the aromatic ring conjugation in the azo dye structure. OR2 and AR18 are hydroxyl azo aqueous species forming hydrazones (o-ortho and/or p-para), while AR 66 is typically an aminoazo dye where azo tautomeric arrangements predominate [41], despite the formation of the o-hydrazone in another molecular region.

Particularly, the OR2 azo dye occurs as an o-hydrazone (Figure 6a) without perturbation of the sulphonate group (-SO₃^-), which is p-substituted in the phenyl (ph), while in AR18 (Figure 6b), the o-hydrazone is protected or hindered by -SO₃^- at positions 2 and 4 of the naphtyl (naph) group adjacent to the azo bond, and there is a third -SO₃^- p-substituted group in the second naphtyl group of the molecule. Thus, the high number of sulphonyl groups in AR18 inhibits by electrostatic repulsion the ^•OH attack and the other reactive oxygen species (ROS) and reduce the steric hindrance and the accessibility the oxidizing radicals (^•OH, ¹O₂, O₂^•-, HO₂^•) to the azo groups and aromatic azo rings [42]. Under these conditions, OR2 is easily degraded by ROS generated on the Fe₃O₄ (≡Fe³⁺) surface catalyst of AR18, as observed in the direct photodegradation of these dyes by UV irradiation in a rotating disk photoreactor [43]. In this direct photolysis, a pseudo-first order constant ratio for Orange 2 and Acid Red 18 (k₁ OR2/k₁ AR18) was 1.97, which is similar (~ 2.17) to that found in this work for photocatalytic degradation on synthesized magnetite. For AR66, the k₁ (min^-1) is between the k₁ for OR2 and AR18, which can be explained in function of the minor steric hindrance and electrostatic repulsion generated from two structural sulphonate situated in position 2 of the o-hydrazone and another p-substituted in the phenyl group associated with the second azo group of the molecule (Figure 6c). Besides, the conjugation of the rings increment in the disulphonate (AR66) type: naph-azo-ph-azo-ph is more favorable to photoconversions than the naph-azo-naph conjugation present in AR18 (trisulphonate).

Based on the hard/soft acid/base (HSAB) of Pearson theory [44]; Fe³⁺ is typically a strong strength Lewis acid (HA) with high affinity for strong strength Lewis (HB), giving ionic interactions for base HA/HB complex interactions. On the other hand, azo dyes can be characterized as soft bases (SB) with a basic character decreasing with increasing carbonic chain length, which generates a diffuse electron distribution. Thereby, the minor molecular complexities (MC) and topological polar surface area (PSA) in the OR2 (MC = 529, PSA = 111 Å) and AR18 (MC = 958, PSA = 201 Å) in comparison with that of AR18 (MC = 1080, PSA = 242 Å) (data obtained from Pubchem databases) can favor the catalytic surface interactions, promoting the heterogeneous mechanistic oxidation. In this order, the polar covalent interaction features in the conjugated Fe³⁺//OR2 complex are stronger than in than the Fe³⁺//AR18 base system, whereas the Fe³⁺//AR66 conjugated par appears as an intermediate character.

When considering the catalyst participation in the processes, it is important to figure out the position and nature of the two active iron species on the surface (≡Fe³⁺ and/or ≡Fe²⁺) in contact with the reactive functional groups (-N=N-, -SO₃^-, -C=C-) of the excited dye molecule (dye*). The heterogeneous photo-Fenton process is described by the following equations (Eq.1 to 6), being the ^•OH generation due to the H₂O₂ direct photolysis at 180-340 nm (photochemistry reaction, Eq. 2) and iron cations (photocatalytic reactions, Eq. 3 and 4) [3,15,45].

dye + hλ → dye*

(1)

H₂O₂ + hλ → 2 ^•OH

(2)

≡Fe³⁺ + H₂O₂ + hλ → ≡Fe²⁺ + ^•OH + H⁺

(3)

≡Fe²⁺ + H₂O₂ → ≡Fe³⁺ + ^•OH + ^-OH

(4)

≡Fe²⁺ + H₂O₂ → ≡Fe³⁺-OH + ^-OH

(5)

dye* (organic) +^•OH → products

(6)

Magnetite synthesized is a magnetic mixture of hematite (α-Fe₂O₃), maghemite (γ-Fe₂O₃) and goethite (α-FeOOH) having about 70% of Fe₃O₄ and 30% of the aforesaid phases [46]. Such composition is non-stoichiometric and is abundant in iron species that act favorably in the ^•OH radical formation. Figure 7 show a comparison of the AR 18 degradation using various iron oxides with different Fe³⁺/Fe²⁺ molar ratios, for instance, wüstite (FeO, Fe³⁺/Fe²⁺= 0), goethite (FeOOH, Fe³⁺/Fe²⁺= ꝏ), hematite (Fe₂O₃, Fe³⁺/Fe²⁺= ꝏ) and magnetite (Fe₃O₄, Fe³⁺/Fe²⁺= 4) prepared in this study.

At 25 ^oC, the initial concentration of 60 mg.L^-1 of the dye is not a limiting factor for the catalytic activity of the various iron oxides (Figure 10), as observed for example at the concentration of 200 mg.L^-1 (0.33 mmol.L^-1) of Acid Red 18, where the maximum degradation was 62.3% (Figure 1b). The similar degradations indicate similar mechanistic considerations for catalytic surface, despite the oxy (hydroxide) nature of goethite, the major Fe³⁺ proportion seem to contribute to a slight higher oxidation to that of FeO, possessing only Fe²⁺ sites. In photo-Fenton systems with high Fe²⁺ concentrations under acidic conditions, Fe²⁺ exists in the Fe(OH)²⁺ form, which determines a lower ^•OH radical generation and the decrease of the degradation rate [49].

2.3. Treated Solution Characterization

Figure 8 shows the UV-Vis spectra of the OR2 azo dye treatment, after the heterogeneous photo-Fenton reaction over the synthesized Fe₃O₄. The structural iron reacts with H₂O₂ giving rise to ^•OH radicals, which attacks the azo (-N=N-) groups producing an effective solution decolorization in 90 min of photocatalytic reaction. The UV decolorization at 230 nm and 310 nm is related to the benzene, naphthalene ring as shown in literature reports, although these bands have a much lower and slower intensity than the chromophore band positioned at 505 nm. These behaviors are similar to those observed in the degradation of the azo dyes Reactive Black 5 [50] and Orange 2 [51], indicating a high decolorization of the OR2 accompanied by a relative concentration of the residual compounds after the oxidative reactions. Regarding the chemical oxygen demand (COD) before and after the photo-Fenton reaction, only a reduction of 29.7% in chemical oxidation of organic matter, which is similar to the COD of 27% found in the degradation of the dye Acid Red 73 by heterogeneous photo-Fenton process [52]. These results reveal the need for long reaction times or harsh treatment conditions aiming the mineralization of the refractory compounds produced.

2.4. Spent Magnetite Characterization

The fresh magnetite characterizations are given in our previous reports [46]. XPS spectra of fresh and spent magnetite after the poto-Fenton-like reaction are shown in Figure.9. The XPS spectrum of fresh Fe₃O₄ (Figure 9a) is depicted with the main spin-orbit split peaks located at 710.2 and 724.9 eV (Table 3) for the doublet Fe 2p_3/2 - Fe_5/2, respectively. These doublet peaks arise along with their shake-up satellites typically confirming that the magnetite has Fe species in both divalent Fe³⁺ and Fe²⁺ states [53]. Accordingly, the peak at about 710.8 eV is assigned to the Fe³⁺ ions in octahedral sites whereas that located at 723.9 eV indicates the presence of the Fe²⁺ species [54].

The recorded O 1s core level has three different oxygen species and a major contribution at 529.6 eV corresponds to 63%, due to the existence of lattice oxygens from the Fe-O bonds in Fe₃O₄,as previously detected in the literature [55]. This contribution is accompained by two other weak components near 531.0 eV (29%) from OH groups of physisorbed water molecules and 532.7 eV (8.0%) agreeing well with carbon bonded to oxygen species e.g., C–O and C=O [46,54].

Moreover, Fe³⁺/Fe²⁺ atomic composition obtained from area ratio between the peak at 711.2 and 710 eV is 2.3 (Table 3) suggests the distribution of Fe³⁺ and Fe²⁺ ions between the different sites of magnetite, which has highly oxidized surface [46]. The C 1s core level appears with binding energies ascribed to –C=C– and adventitious carbon as the dominant peak at 284.7 eV (82%) along with peaks at C–OH at 286.0 eV (12%) and C–C=O– at 288.3 eV (6%), in agreement with the findings [53,54,55,56].

These species arises from the incomplete organic precursors removal, even after the calcination of the solid.

In particular, the S 2p core level appears only for Fe₃O₄/R66, having binding energy at 168.1 eV due to S-O groups [46]. Additionally, the detected sulfates species on the surface of magnetite after AR66 Fenton-like reaction would contribute to the physisorption of the Fe active sites on magnetite providing a low performance of the solid with using the AR66 dye.

The atomic concentration percentages of Fe, C and O in fresh Fe₃O₄ are 29.3, 56.5 and 14.4%. Comparison with the atomic concentrations of the components of the two dyes show values of C: 41.5%, O: 45.7%, Fe: 12.6% and C: 37.5%, O: 51.6%, Fe: 10.8%, S: 0.18% for Fe₃O₄/AR18 and Fe₃O₄/AR66 with deviations within the limits of precision and accuracy of the XPS technique.

Besides, Table 3 indicates that the Fe³⁺/Fe²⁺ atomic ratio involves a 2.4-2.6 range for both dyes, which might be expected, according to the surface composition of magnetite [57]. It suggests the magnetite stability towards the reaction.

3. Materials and Methods

3.1. Materials

Pure Acid Red 18 (C₂₀H₁₁N₂Na₃O₁₀S₃, 604.48 g.mol^-1) azo dye was purchased from Acros Organics. Orange 2 (C₁₆H₁₁N₂NaO₄S, 85%, 350.32 g.mol^-1) and Acid Red 66 (C₂₂H₁₄N₄Na₂O₇S₂, 55%, 566.48 g.mol^-1) azo dyes were obtained from Sigma-Aldrich. All azo dyes reagents were used without further purification. The hydrogen peroxide solution (H₂O₂, 30 wt. %) was purchased from Êxodo Científica. Prior to the use, the content of H₂O₂ was evaluated by permanganate method [58]. The sulfuric acid (H₂SO_4, 98%) was obtained from Vetec. Deionized water was used in all experiments.

3.2. Catalyst Synhesis and Characterizations

Fe₃O₄ consisting primarily of a Fe³⁺/Fe²⁺ molar ratio of 4 and was synthesized by a coprecipitation method, as found elsewhere [46,59]. The zeta potential and the particle size distributions at pH of 3.0, 6.0 and 8.0 were performed in a Zetasizer Nano S apparatus form Malvern.

The surface compositions and valence state of the elements present in the spent samples used at distinct reaction conditions were determined by X-ray photoelectron spectroscopy (XPS). XPS spectra were obtained on a PHI Versa-Probe II Scanning XPS Microprobe from Physical Electronics. The incident radiation was generated by the Al Kα X-ray source, operating at 1486.6 eV and 15 kV. The spectra were calibrated using the binding energy peak of C 1s at 284.8 eV from adventitious carbon.

3.3. Heterogeneous Photo-Fenton-Like Oxidation

The photocatalytic activity of the magnetite was conducted in a borosilicate reactor (height of ca. 13 cm; diameter of 6.5 cm) with a capacity of about 45 mL coupled to a circulating water jacket to maintain solution temperature. A UV-C Philips TUV PL-S 7W/2P model lamp of ca. 7.1 W was positioned to the reaction reactor. The aforesaid lamp generated UV irradiation at 254 nm with a radiation intensity (I_L) of ca. 16.7 mW.cm^-2.

Typically for photo-Fenton-like batch experiments, the reaction was initiated by introducing the H₂O₂ solution along of the azo dyes to the reaction system. Then, air flow at a 180 ± 20 mL.min^-1 rate was introduced in the system, which was used to have a homogeneous solution. The abovementioned suspension was exposed to UV-C lamp and aliquots were withdrawn to determine the concentration of the azo dyes solution by a UV-Vis spectrophotometer. When using Acid Red 18 (AR18) as reactant, the pH, initial concentration of H₂O₂ and an initial concentration of the AR18 azo dye were evaluated at 25 ^oC. Other azo dyes such as AR66 and OR2 having an initial concentration of ca. 0.33 mmol.L^-1 were also examined. The calibration curve of each azo dye was set at 505, 510 and 485 nm for AR18, AR66 and OR2, respectively, using a Thermo Fisher UV-Vis Genesys 10 S spectrophotometer. Iron concentrations in solution were estimated by atomic absorption spectrophotometry using a Thermo Scientific AAS iCE 3000 equipment.

The effects of initial concentration azo dye, H₂O₂ initial concentration, catalyst dosage, pH besides kinetic and thermodynamic parameters were evaluated. Besides, initial dosage of catalyst was taken from literature reports [51,60].

The photo-Fenton catalytic degradation percentage was measured, as follows (Eq. 7):

$$\mathrm{\%}\mathrm{D}\mathrm{e}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=\left({\displaystyle \frac{{\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{t}}}{{\mathrm{C}}_0}}\right)*100$$

(7)

where: C₀ (mmol.L^-1) and C_t (mmol.L^-1) are the initial and final azo dyes concentrations at a given period, respectively.

To understand the Acid Red 18 azo dye at 25, 40 and 60 °C degradation kinetics, the pseudo-first-order kinetic model were used, according to the equation 8 [34]:

$$C=C_0{e}^{-k_{1}t}$$

(8)

where: C₀ is the initial concentration of AR18 (mmol.L^-1), C is the AR18 concentration (mmol.L^-1) at a certain reaction time t (min) and k₁ (min^-1) is the rate constant for pseudo-first-order reaction.

The activation energy (E_a, kJ.mol^-1) of the AR18, AR66 and OR2 l studied azo dyes was determined by the Arrhenius equation (Eq. 9), as reported in a previous [40]:

$$k_1=\mathrm{A}{\mathrm{e}}^{{\displaystyle \frac{-Ea}{RT}}}$$

(9)

where: A is the Arrhenius constant, k₁ is the rate constant for pseudo-first-order reaction, T is the absolute (K) and R is the universal constant of the gases (8.314 J.mol^-1.K^-1).

The variations of free Gibbs energy (ΔG), entropy (ΔS) and enthalphy (ΔH) themodynamics were determined according to the van’t Hoff equations [61].

$$k_e={\displaystyle \frac{C_{{\mathrm{A}}_{\mathrm{e}}}}{C_{{\mathrm{S}}_{\mathrm{e}}}}}$$

(10)

$$\mathsf{\Delta }\mathrm{G}°=-RT\ln k_e$$

(11)

$$\ln k_e={\displaystyle \frac{\mathsf{\Delta }\mathrm{S}°}{R}}-{\displaystyle \frac{\mathsf{\Delta }\mathrm{H}°}{RT}}$$

(12)

where: k_e is the equilibrium constant and C_Ae and C_Se are the degraded and remained azo dyes concentrations in the equilibrium, respectively.

4. Conclusions

The synthesized magnetite showed high photocatalytic efficiency for the degradation of dyes AR18, AR66 and OR2 under optimal conditions of 1.4 g.L^-1 of catalyst, 60 mg.L^-1 of H₂O₂, pH = 3.0 and radiation intensity of 16.7 mW.cm^-2. The material showed no decrease in catalytic activity after six consecutive reaction cycles, confirming its high stability, as observed by the XPS results. The kinetics were typically first order while the reactional thermodynamics confirmed spontaneous and endothermic degradation processes. Among the dyes, the OR2 molecule was the most easily degraded due to its lower structural complexity, lower steric hindrances and lower electrostatic repulsion with reactive oxygen species on the catalyst surface. This study showed the promising nature of photo-Fenton technology using magnetite as a catalyst in the treatment of aqueous solutions contaminated with azo dyes, also allowing to discuss and explain the effect of some chemical characteristics of the target molecule on the performance of oxidative degradation.