Enhanced degradation of 2,6-dimethylphenol by photocatalytic systems using TiO2 assisted with H2O2 and Fe(III)

ABSTRACT In this study, several photocatalytic degradation systems were investigated using 2,6-dimethylphenol (2,6-DMP) as a model compound. Highly reactive species are formed in four systems, Fe(III), TiO2, TiO2/H2O2 and TiO2/Fe(III) where complete degradation of 2,6-DMP was achieved under UV radiation. Photodegradation of the 2,6-DMP has been described by pseudo-first order kinetic model in the presence of TiO2. In UV/TiO2–H2O2 system, the addition of H2O2 in the TiO2 suspension improves the degradation rate of 2,6-DMP from 70% to 100% for a H2O2 concentration of 10− 2 M in 3 h. In homogeneous system, HO• and Fe2+ can be generated by the irradiation of Fe(III) solution. The speciation of Fe(III) obtained from Visual MINTEQ soft showed the formation of several species and Fe(OH)2+ were the most predominant and active species in a pH range of 2.5–3.5. At a low concentration of TiO2 (30 mg L−1), an important positive effect due to the iron addition has been shown in TiO2/Fe(III) system, the entrance of metallic ions at different concentrations enhanced the photocatalytic activity of TiO2. A degradation percentage of 90% was achieved in the UV/TiO2–Fe(III) system under optimal conditions against 57% in UV/TiO2 system. Strong synergistic effect was observed in the UV/TiO2–H2O2 binary system. On the basis of literature, a pathway for 2,6-DMP degradation was proposed. The mechanism of degradation of the 2,6-DMP did not involve only HO• radicals, an interaction of Fe(III) in the excited state with 2,6-DMP occurred giving rise to the formation of 2,6-dimethylphenoxyl radical. GRAPHICAL ABSTRACT


Introduction
Removing and mineralization of phenolic compounds from water and wastewater is a big challenge, which can be achieved through several treatment processes, including biodegradation [1,2], advanced oxidation processes [3,4], membrane filtration [5], etc. These hazardous organic pollutants are generally existing in petrochemicals, pesticides, pharmaceuticals, paints, plastics, oil refinery and paper manufacturing industries [6][7][8]. Among the various organic pollutants treatment methods, advanced oxidation processes (AOPs) involving the in-situ generation of highly oxidative hydroxyl radicals ( • OH, E 0 = 2.8 V vs. NHE) has been reported as an efficient method for the degradation and mineralization of aqueous pollutants [9,10]. The most commonly used AOPs include heterogeneous photocatalysis in the presence of titanium dioxide (TiO 2 ) which is widely used in the pollution of air and water of a large number of organic pollutants, since its first application in 1972 [11] and until now [12][13][14]. TiO 2 , as a versatile material, has been considered as a well-known photocatalyst because of its photochemical stability, low toxicity, good chemical/physical stability and low cost [15,16]. The principle of this technique is well known, it depends on the activation of TiO 2 by providing light energy (λ < 400 nm). At this stage the activation has created a pair of electron/hole (e − /h + ), which follows the passage of electron from the valence band to the conduction band, creating an oxidation site (h + ) and a reduction site (e − ) (reaction (1)). Then, after separation, the excited holes and electrons act as oxidizing and reducing agents, respectively, to generate hydroxyl radicals ( • OH) (reactions (2)-(6)).
Recombination of h + /e − is considered as a disadvantage or a reaction limiting the efficiency of photocatalysis using suspensions of titanium dioxide [17]. Over the past 10 years, the research has been concentrated to reduce the effect of recombination of charges or to increase the performance of photocatalysis using different techniques. Among them, adding an electron acceptor such as oxygen that is naturally present in water, ozone [18], hydrogen peroxide (H 2 O 2 ) [19], H 2 O 2 /O 3 [20], sulphite or peroxymonosulphate [21,22] and other substances such as transition metals [23,24]. Similar to Fe(III), ferrate(VI) could be also excited by UV irradiation to produce highly reactive species or acted as capture agent of the electrons from photocatalytic systems [25]. Modification of TiO 2 surface was also used to enhance the degradation of organic pollutants [26]. Doping of TiO 2 , by noble metals [27], nonmetal [28] or transition metals, frequently by Fe (III) [29] changes its structure and/or its morphology and consequently can modify the photocatalytic activity. Synthesis methods are sometimes very complex and the reaction yield is not always high, which can increase the cost of catalyst production [30]. Rengifo-Herrera et al. [31] reported that the new synthesized materials, N, S co-doped and N-doped commercial anatase TiO 2 powders did not show improvement for the photocatalytic degradation of phenol and inactivation of E. coli under simulated sunlight. They found the undoped Degussa P-25, commercial powder, exhibited the highest photocatalytic activity. However, adding H 2 O 2 or Fe (III) ions to TiO 2 suspensions will be a simple, efficient and inexpensive process. Irradiation of TiO 2 suspensions generates H 2 O 2 (reaction (6)) but in insignificant amounts as recently demonstrated by Zhang et al. in 2020 [32]. However added H 2 O 2 can provide an additional source of • OH radicals by capturing the electron from the valence band as shown in reaction (7) which enhance considerably the degradation rate of organic pollutants [33][34][35][36].
In aqueous solution, Fe (III) is used alone as photoinducteur in order to degrade different organic substrates [37,38]. Recently, Sadhu et al. [39] studied the degradation of Reactive Black 5 (RB5) using Fe(III) ions under solar light, more than 90% RB5 was decolourized in 15 min. All Fe(III) species undergo a photoredox reaction under light irradiation [40,41]. Fe(III) ions are photoreduced by ligand-to-metal charge transfer (LMCT) giving rise to Fe 2+ and HO • radical as shown in reaction (8) below: The addition of Fe 3+ ions to TiO 2 aqueous solution can increase the efficiency of organic compounds degradation during UVA irradiation [42][43][44], by trapping the electron photogenerated by TiO 2 particles as shown in reaction (9). An additional probable reason is the photochemical synergy of TiO 2 and (Fe(III)-OH) species under UV irradiation [45].
In ternary combined system, UV/TiO 2 /Fe(III)/H 2 O 2 , in addition of photocatalytic effect of TiO 2 , dual roles of Fe (III) synergistically boosts • OH production at neutral pH [46]. First Fe(III) acts as an electron acceptor preventing the recombination of electron-hole pairs of photocatalyst. Second, resulting Fe(II) ions transform H 2 O 2 into • OH giving rise to Fenton and photo-Fenton processes [47]. Recent researches studied the influence of adding H 2 O 2 and/or Fe(III) into TiO 2 suspensions and reported that these green oxidants can greatly influence the photocatalytic system for the generation of additional • OH radicals. Zulfiqar et al. [48,49] in two different studies reported that the introduction of FeCl 3 or H 2 O 2 in the photocatalysis and adsorption systems in the presence of TiO 2 nanosheets (TNSs) and TiO 2 nanotubes (TNTs) produced additional • OH and O †− 2 radicals that would be more responsible to improve the removal of phenol. Another recent study by poulopoulos et al. [50] on the photocatalytic treatment of phenolic compounds as, phenol, 2-chlorophenol, 2,4-discholophenol, trichlorophenol, and 4-nitrophenol in a synthetic wastewater was investigated in an annular photoreactor under UV irradiation. The results showed the total carbon (TC) removal obtained for TiO 2 /H 2 O 2 /Fe(III)/UV system was 84% vs. 58% in TiO 2 /UV system. 2,6-dimethyphenol (2,6-DMP), used as a model pollutant, is one of the six isomers of xylenols which is an essential element of phenolic derivatives [3]. In a previous study in our laboratory, its degradation was followed in natural iron oxide (NIO) and oxalate system under UV irradiation based on hydroxyl radical generation and Fe (III)-oxalate complexes speciation [3,51,52]. In this study we investigated the photocatalytic degradation of 2,6-DMP using four oxidation systems TiO 2 , Fe (III), TiO 2 / H 2 O 2 and TiO 2 /Fe(III). At first, we studied the photocatalytic behaviour of 2,6-DMP in presence of TiO 2 and the influence of the addition of H 2 O 2 in the suspensions. Then the degradation of 2,6-DMP in homogeneous iron (III) solutions and in the combined system TiO 2 /Fe (III) was investigated under UV irradiation. Moreover, the evaluation of the influence of some parameters such as 2,6-DMP concentration, Fe(III) concentration and the determination of the main iron species and radicals involved in the degradation of 2,6-DMP are investigated. It should be noted that 2,6-DMP does not form a complex with Fe(III) and it is among the few components which present an oxidation in the presence of Fe(III) in the dark. This property constitutes an advantage, which allows the degradation of this component without having the need to UV or visible light however; this will reduce the cost of the treatment and this might be a subject for further study.

Irradiation procedure
The photocatalytic experiments were performed in a self-constructed Pyrex glass tube with a double envelope placed in a cylindrical aluminium container as described in previous work [54]. A Philips TLAD ultraviolet lamp with the following specifications: nominal wattage 15W/05, wavelength 365 nm (dominantly), diameter 30.00 mm and length 560 mm was used. Its light intensity (I = 0.45 mW cm −2 ) was measured using a radiometer type VLX3W. All experiments were performed at T = 20 ± 2°C by cycling water. The irradiated solutions and suspensions were magnetically stirred during whole experiments. Before irradiation, 2,6-DMP and TiO 2 suspensions were stirred in the dark for 30 min to establish adsorption/desorption equilibrium. The particles of TiO 2 were removed after irradiation by filtration through cellulose acetate membrane (Millipore 0.45 μm). The oxidation experiments with Fe(III) were performed in the same conditions without switching on the UV lamp. All aqueous solutions were prepared with ultrapure water obtained from a Milli-Q ultrapure water purification system (Millipore, Bedford, MA, U.S.A.).

Analytical techniques
The solutions absorption spectra have been recorded on a spectrophotometer Unicam 'Heλios α'. The disappearance of 2,6-DMP was quantified by high-performance liquid chromatography using a Shimadzu HPLC equipped with a controller model SCL-10A VP, photodiode-array UV-Vis detector model SPD-M10A VP and two pumps model LC 8A. The system is controlled by software 'Class VP5' for storing and processing of chromatograms. The analytical column is a C18 Supelco (5 μm, 250 mm × 4.6 mm i.d). The eluent was a mixture of methanol/water (50/50 v/v) at a flow-rate 1 mL min −1 . The detection wavelength was at 270 nm.
The monomeric species (Fe(III)-OH) concentration of Fe(III) was determined by complexometry with 8-hydroxyquinoline-5-sulfonic acid (HQSA) [55]. Fe(III)-OH refers to Fe(III), Fe(OH) 2+ and Fe(OH) 2 + monomeric species present at pH ≤ 5. HQSA reacts much more rapidly with Fe (III) monomeric species than with Fe (III) dimer or Fe (III) polymers [40,55]. The absorbance of the complex Fe(HQSA) 3 was measured at 572 nm, the molar absorption coefficient determined and used in this study was equal to 5000 ± 100 M −1 cm −1 . This value agrees well with the value reported in the literature [40,55]. The percentage of monomeric species Fe (III)-OH was determined as the ratio of [Fe(III)-OH] to total [Fe (III)] in solution by the following equation: where [Fe(III)]tot is the total initial concentration of Fe (III).
The concentration of Fe (II) was determined by complexometry with 1,10-phenanthroline. The molar absorption coefficient, for the Fe(II)-phenanthroline complex, at 510 nm was 11,180 M −1 cm −1 [3]. It was carefully checked that no interference in the analysis was observed when 2,6-DMP was present in the solution.

Results and discussion
3.1. Photocatalytic degradation of 2,6-DMP in presence of TiO 2 2,6-DMP degradation was studied in two different experimental conditions: (i) under UV illumination in the absence of TiO 2 and (ii) under UV illumination in the presence of TiO 2 (Degussa P25) ( Figure 1). As can be seen, irradiation in the absence of P25 showed no significant degradation. However, irradiation of aqueous solution of 2,6-DMP (5 × 10 −4 M) in the presence of TiO 2 (1 g L −1 ) at experimental pH = 5.8, led to the complete degradation of the substrate. The oxidation of 2,6-DMP in the presence of TiO 2 follows pseudo-first-order kinetic (inset in Figure 1), The rate constant calculated was 0.44 h −1 (R 2 = 0.977). The disappearance of 2,6-DMP is quite fast at the beginning of the reaction with a half-life time equal to 81 min. For long irradiation times, the increase in the amount of generated intermediates alters the kinetic and decreases the rate of oxidation of 2,6-DMP. They create a competition for adsorption sites of the semiconductor and so that for reactive photogenerated species. The total disappearance of 2,6-DMP was obtained after 8 h of irradiation.
3.1.1. Effect of initial concentration of 2,6-DMP Photocatalytic degradation of different initial concentrations of 2,6-DMP in the range of 10 −4 -10 −3 M were assisted by 1 g L −1 of TiO 2 particles. The results are shown in Figure 2. The rate degradation of 2,6-DMP decreases with the increase in initial concentration from 10 −4 to 10 −3 M. The results showed that the time needed for the removal of 2,6-DMP was found to depend considerably on the initial substrate concentration. Complete degradation of 2,6-DMP occurred in 3 and 5 h for relatively low initial concentration values of 10 −4 and 2 × 10 −4 M respectively while for higher initial concentration, 5 × 10 −4 and 10 −3 M, a complete degradation occurred and a rate of degradation of 90% was registered in 8 h respectively.
The decrease in degradation effectiveness with the increase in substrate concentration happens due to several reasons. At high pollutant concentration, the adsorbed reactant molecules lead to the saturation of the surface of titanium dioxide and significant amount of UV is absorbed by the organic compound molecules rather than the TiO 2 particles [56,57]. Hence, a decrease in the absorption of photons by the photocatalyst particles occurs, which obviously leads to a decrease in the amount of reactive species ( • OH and O 2
2,6-DMP photocatalytic degradation can be considered as a pseudo-first order chemical reaction (inset in Figure 2). The kinetic parameters are summarized in Table 1.  respectively. It can be found that the rate constants increase with an increase in the H 2 O 2 concentration. The UV/TiO 2 -H 2 O 2 system exhibited weak photoreactivity for [H 2 O 2 ] from 10 −4 to 10 −3 M and for [H 2 O 2 ] from 10 −2 to 5 × 10 −2 M, it exhibited much higher photoreactivity. The results describe that the system UV/TiO 2 -H 2 O 2 is more effective than photocatalyst alone. This conclusion has been achieved in a large number of previous studies, for example Moura et al. [61] and Monteagudo et al. [62] which showed that the addition of hydrogen peroxide to titanium dioxide increases the rate of disappearance of a mixture of the non-steroidal anti-inflammatory drugs ketoprofen, meloxicam and tenoxicam and the aniline respectively. As explained above, the added H 2 O 2 accelerates the reaction by scavenging electrons of the conduction band and producing additional hydroxyl radicals depending on the reaction (7). This reaction has a dual effect: (1) it provides a photocatalytic process, an additional pathway for the production of hydroxyl radical to accelerate the degradation efficiency of intermediate compounds and (2) removing the electron-hole recombination by accepting the conduction band electron.
However, the initial reaction rates show a slight change in the range of concentration of H 2 O 2 10 −2 −5 × 10 −2 M. This phenomenon can be explained by the reaction of self-inhibition following [63]: H 2 O 2 plays a dual role: it contributes to the formation of hydroxyl radicals by the reaction (7) and is involved in their capture by the reaction (11). At low concentrations, the first reaction is predominant and there is an acceleration of the photocatalytic degradation. The gradual increase in the concentration of H 2 O 2 permit, however, the reaction (11) to take an important growing. ) and finally, ferric oxides and hydroxides are also present in the form of suspended particles such as: α-FeOOH (goethite), α-Fe 2 O 3 (haematite), etc. The species distribution as a function of pH is based on the hydrolysis equilibrium [39,64,65] shown as reactions (S1-S5)) in SI data. The speciation of Fe(III) under the test conditions (298 K with ionic strength 0.03 M) was analysed using Visual MINTEQ (version 3.0) soft [66,67], the modelled species include seven complexes at different concentrations of Fe(III) (Figure 4(A)) and (Figure S1-S3) in SI data. In fact, at pH ≤ 5, five different species, Fe 3+ , Fe(OH) 2+ , Fe(OH) + 2 , the dimer Fe 2 (OH) 4+ 2 and the trimer Fe 3 (OH) 5+ 4 coexist in aqueous solution in different percentages (formation of iron precipitation was omitted) (Figure 4(B)), obviously, Fe(OH) 2+ is the predominant species at this range of pH. Fe(III) speciation was investigated in the range of Fe(III) concentration from 5 × 10 −4 to 6 × 10 −3 M. The results are gathered in Table 2. From the values of the quantum yields, we can see that Fe(OH) 2+ and Fe(OH) + 2 monomer complexes are the most photoactive species for the formation of hydroxyl radicals. The other two species, Fe 3+ and Fe 2 (OH) 4+ 2 ) are also sources of hydroxyl radicals under irradiation but have much lower quantum yields for approximately the same wavelength.

Behaviour of 2,6-DMP-Fe(III) system in the dark
The absorption spectra of (2,6-DMP-Fe(III)) (5 × 10 −4 -5 × 10 −4 M) mixture with 98% of monomers complexes in the dark and at room temperature was recorded. As shown in Figure 5, three distinct bands are recorded, the absorption band at 297 nm corresponds to Fe(OH) 2+ species (as shown in Figure S4 for freshly prepared solution) [39], Ageing of Fe (III) solution in the dark and at room temperature corresponds to a decrease of the iron (III) monomer complexes and to the formation of dimeric or oligomeric species. At the same time, the determination of the monomer complexes by complexometry with 8-hydroxyquinoline-sulfonic acid (HQSA) in the same solution of Fe(III) clearly shows the disappearance of these species at different times after preparation (inset in FigureS4). The band at 270 nm corresponds to 2,6-DMP and another band at 420 nm which can be related to the formation of diphenoquinones as intermediates compounds [70][71][72].
In dark, an oxidation of 2,6-DMP took place during the addition of iron salts in the solution. It was characterized by a change in appearance of the solution (yellowish colour appeared). It is noted that the absorbance at 270 nm increases because of the appearance of intermediates compounds such as hydroquinone and benzoquinone which absorb at wavelengths close to that of 2,6-DMP.
In Figure 6(A), the quantification of 2,6-DMP by HPLC shows clearly a slight degradation in presence of two concentrations of Fe(III), 5 × 10 −4 and 8 × 10 −4 M. Degradation rates of 23% and 34% for 5 × 10 −4 and 8 × 10 −4 M in 3 h respectively were recorded. The results showed that the degradation of 2,6-DMP depends on the concentration of Fe(III). At the beginning, the degradation of 2,6-DMP was moderately fast, slowed down and   reached a plateau value around 2 h for the two concentrations. As explained in the literature, the process was an electron transfer oxidation giving rise to Fe(II), the reduced form of the predominant specie Fe(OH) 2+ aqua complex and 2,6-dimethylphenoxyl radical as described in reaction (12): (12) Bacon et al. [70], Cecil et al. [71] and Mazelier et al. [72] investigated the oxidation of 2,6-DMP by metal ions and complexes. They suggested that 2,6-DMP might be easily oxidised by a simple electron transfer in systems where the oxidant [hexachloroiridate (IV) anion or Fe (OH) 2+ cation …] is a labile complex but the reductant is an organic compound (2,6-DMP) without marked tendencies to become a ligand. They also proposed that dehydrogenative coupling reactions are the main ones detected in oxidation of phenols giving dimers or large polymers and the oxidation occurring in ortho or para positions. Effectively, Belattar et al. [73] confirm the absence of an interaction between Fe(III) species and 3,5-DMP (meta positions).
During these experiments, the formation of Fe(II) was followed by complexometry. Similarly for 2,6-DMP disappearance, the Fe(II) concentration immediately increased and reached a plateau value around 2 h for the two concentrations ( Figure 6(B)). As it can be seen, the Fe(II) concentration increased with increasing Fe(III) concentration. Consequently, the formation of Fe(II) confirm the involvement of monomeric species in a redox reaction with 2,6-DMP in the dark and at room temperature.

Degradation of 2,6-DMP photoinduced by Fe (III)
To carry out the photochemical experiments, we placed in conditions where the contribution of the oxidation in the dark process is at the lowest possible level. Thus irradiations were carried out immediately after mixing 2,6-DMP and iron (III) and the samples were analysed immediately after sampling. The pH of the solutions is equal to 3.  (Figure 7(B)). Higher Fe(III) concentration led to a higher degradation of 2,6-DMP and higher concentration of Fe(II). The rate of degradation of 2,6-DMP increased approximately to 40% and 60% in 3 h for Fe(III) concentrations of 5 × 10 −4 and 8 × 10 −4 M respectively. The formation of Fe(II) was very fast at the beginning and reached a limit value of about 4.95 × 10 −4 and 7.01 × 10 −4 M for Fe(III) concentrations of 5 × 10 −4 and 8 × 10 −4 M respectively which is approximately equal to that of the initial concentration of Fe(III). Similar results were reported by a number of authors [38,74,75]. Under UV light irradiation, Fe (III) hydroxylcomplexes undergo photochemical reduction to Fe(II) and formation of • OH radicals can be performed depending on reaction (8).  to better show the effect of iron ions in TiO 2 suspension. Mestankova et al. [45] reported that the positive effect of added iron ions has clearly observed for lower concentrations of TiO 2 . The initial concentration of (Fe(III)-OH) species was kept constant for both experiments in the presence of iron and equal to 98%.
The results showed that the degradation of 2,6-DMP in the combined system (TiO 2 -Fe(III)) was faster than by TiO 2 alone and Fe(III) alone, indicating that the degradation rate of 2,6-DMP could be enhanced by the addition of Fe ions in TiO 2 suspension. About 78% of 2,6-DMP was degraded by TiO 2 -Fe(III) photocatalysis after 7 h irradiation, however 47% and 66% of 2,6-DMP were degraded by Fe(III) and TiO 2 after 7 h irradiation respectively. These rates indicate the photoproduction yield of hydroxyl radical in UV/TiO 2 -Fe(III) system is higher than that in other two systems. The positive effect of added Fe(III) cations in TiO 2 suspension can be attributed to the electric trapping at the semiconductor surface [48,49,76], in the combined system Fe(III) reduces the recombination rate of electron and holes in TiO 2 (P25) by the reaction (9) cited above.

Effect of initial Fe(III) concentration in UV/ TiO 2 -Fe(III) system
To better show the effect of iron addition on 2,6-DMP degradation in Fe(III)-TiO 2 combined system under irradiation at 365 nm, a set of experiments were carried out ( Figure 9). The concentration of Fe(III) was varied from 5 × 10 −4 to 6 × 10 −3 M, the initial concentration of 2,6-DMP was equal to 5 × 10 −4 M and an amount of TiO 2 was taken as 30 mg L −1 . The addition of ferric ions leads to an increase in the rate constants throughout the concentration range in comparison with suspension without iron. The influence of TiO 2 is evident in the combined system there were no deceleration unlike what was observed in the kinetics with Fe(III) alone ( Figure  S5). This phenomenon is suppressed due to the efficient oxidation of Fe(II) by oxidative species formed upon irradiation of TiO 2 particles (HO • , HO 2 • , …) [44].   Increasing Fe(III) concentration will lead to increased formation of photoactive Fe(III) species, greater electron trapping, greater • OH generation, and therefore greater 2,6-DMP degradation. At pH = 2.5 for 6 × 10 −3 M Fe(III) concentration, the percentage of Fe 3+ (35.4%) and Fe 2 (OH) 4+ 2 (13.8%) species increased, the percentage of Fe(OH) 2+ (47.2%) decreased (Table 2) and the degradation of 2,6-DMP reached 90% in 5 h. So, Fe 3+ species might be the significant contributor for the generation of • OH in the system and that Fe(OH) 2+ species are not the only photoactive species. According to the results of Table 2, an increase in pH value especially in the solutions with pH ≥ 3, show that the concentration of Fe 3 + , Fe 2 (OH) 4+ 2 species decreased, then their contribution to the generation of • OH decreased and can be completely ignored [40,41]. The main species responsible for the degradation of 2,6-DMP in this range of pH and concentrations are Fe(OH) 2+ and Fe(OH) + 2 with lower contribution.

Influence of (Fe(III)-OH) concentration in UV/ TiO 2 -Fe(III) system
Fe(III)-OH monomeric complex is one of the most important parameter for the chemistry and photochemistry of Fe(III) in aqueous solution [77]. To confirm their involvement in 2,6-DMP degradation, the effect of monomeric species percentage was investigated at constant Fe(III) total concentration of 5 × 10 −4 M with percentage values of 98%, 60% and 25% determined by HQSA method. 2,6-DMP and TiO 2 initial concentrations were equal to 5 × 10 −4 and 30 mg L −1 respectively. The monomeric species concentration of 98% was determined for a freshly prepared Fe(III) solution. However, the concentrations of 60% and 25% were determined by leaving the solutions for several days in dark since the ageing of the solution leads to the variation of the chemical composition, formation of dimers, e.g. Fe 2 (OH) 2 4+ and soluble aggregates of Fe(III) [40]. The results are presented in Figure 10(A). The highest degradation rate of 2,6-DMP was observed in samples containing 98% of monomeric complexes. However, a significant slowing down of reactions was observed for lower concentrations (60%) and this effect was more pronounced when the concentration was decreasing to 25%. These results clearly showed that the degradation of 2,6-DMP is mainly caused by the excitation of monomeric complexes and specially by Fe(OH) 2+ predominant species. Other Fe(III) species, such as oligomeric species, appeared to be less photoreactive and their involvement in the degradation process was negligible. Similar results have been described in previous studies [42][43][44]78].  Table 3.
In order to know the main oxidising agent, a mixture of 2,6-DMP/Fe(III)-TiO 2 (5 × 10 −4 M / 5 × 10 −4 M -30 mg L −1 ) with 98% of monomeric complexes was irradiated at 365 nm in the presence of tertio-butanol (tBuOH), which act as a scavenger of hydroxyl radicals. Figure 11 showed the kinetics of the disappearance of 2,6-DMP in the absence and in the presence of 2% of tBuOH. As it can be seen, the disappearance of 2,6-DMP in the suspension of the combined system was not completely inhibited (another amount of tBuOH (3%) was tested and it gave the same kinetic). This experience showed that only a percentage of 33% 2,6-DMP degradation was caused by OH radicals. The inset in Figure 11, shows that the only radicals responsible for the degradation of 2,6-DMP in the presence of TiO 2 alone are OH radicals. On the other hand, as shown in Section 3.2.2, an electron transfer oxidation between 2,6-DMP and Fe(III) leads to the formation of Fe(II) and 2,6-dimethylphenoxyl radical. Therefore, the proposed mechanism of enhanced photocatalysis of TiO 2 by Fe(III) in the present study is that both HO • and electron transfer between 2,6-DMP and Fe(III) are responsible of 2,6-DMP degradation in UV/TiO 2 -Fe(III) system.

Synergistic effect in the two binary systems
The main parameter in the combined processes to assess the efficiency of the system is the synergistic effect (SF). Figure S6 (A) and S6 (B) compare the kinetics degradation of 2,6-DMP obtained from the UV/ TiO 2 -H 2 O 2 and the UV/TiO 2 -Fe(III) binary systems. The synergistic effect could be estimated through the two binary combined systems as follows in Equations (13) and (14) [79,80]: SF = k UV/TiO 2 −Fe(III) k UV/TiO 2 + k UV/Fe(III) (14) where k UV/ As can be seen in Table 4, the value of k UV/TiO 2 −H 2 O 2 is higher than the sum of the value of (k UV/TiO 2 + k UV/H 2 O 2 ) and the calculated synergy factor is equal to 1.  Figure 12. The degradation time taken is 3 h. The results showed that 5% 2,6-DMP degradation was achieved under photolysis while 10% 2,6-DMP   Here, it is necessary to explain the necessity of both oxidizing agents with TiO 2 under heterogeneous photocatalysis process. It should be noted that Fe(III) cations are employed with low concentration of TiO 2 to better show their positive effect. TiO 2 (Degussa P25) is one of the most efficient photocatalysts even when used alone but the presence of Fe (III) and H 2 O 2 plays a significant role in the photocatalytic remediation system by serving as an excellent electron acceptor. They have a relatively strong capacity to produce additional OH radicals which could be useful in breaking the bond of 2,6-DMP molecules and will increase the photocatalytic activity of the photocatalyst. A similar effect has been described by Zulfikar et al. [48,49], Rincon et al. [43] and Mestankova et al. [44,45].

Pathway formation of intermediates in Fe (III)-TiO 2 system
In the current study, intermediates substances are highlighted by referring to previous research. The objective is to show 2,6-DMP degradation pathway in the UV/TiO 2 -Fe(III) system by exploring reaction intermediates generated in the process. Therefore, no identification was made of intermediates.

Conclusion
Removal of 2,6-DMP has been studied in different systems where hydroxyl radicals, highly oxidizing species are mainly formed. In heterogeneous system, complete degradation of 2,6-DMP was achieved by treating the organic compound samples with UV radiation in the presence of TiO 2 (Degussa P25) photocatalyst. In homogeneous system, 2,6-DMP could be efficiently degraded in dark and under UV light in the presence of ferric ions. Photoreduction of Fe(III) monomeric species producing hydroxyl radicals was the main mechanism responsible for 2,6-DMP removal under irradiation and an electron transfer oxidation giving rise to Fe(II) and 2,6-dimethylphenoxyl radical occurred in dark condition. The speciation of Fe(III) was analysed using Visual MINTEQ soft, five different species, Fe 3+ ,-Fe(OH) 2+ , Fe(OH) + 2 , the dimer Fe 2 (OH) 4+ 2 and the trimer Fe 3 (OH) 5+ 4 coexist in different percentages under our conditions (pH = 2.5-3.5). Fe(OH) 2+ is the predominant species at this range of pH. The successful use of the two well-known green oxidants, Fe(III) aqua complexes and H 2 O 2 , as an efficient electron acceptors in the presence of TiO 2 particles under UV irradiation was reported. The results showed that both added H 2 O 2 and Fe(III) can provide an additional source of • OH radicals by trapping the electron photogenerated by TiO 2 particles and minimize the recombination rate of photogenerated hole and electron pairs. The synergetic effect in the UV/TiO 2 -H 2 O 2 binary system was significant for the removal of organic pollutant. In UV/TiO 2 -Fe(III) system, the degradation of 2,6-DMP mainly involves two pathways: the first, the formation of HO • radicals derived from the excited species of Fe(III) and also from excited TiO 2 and the second an interaction of Fe (III) in the excited state with 2,6-DMP.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The author(s) reported there is no funding associated with the work featured in this article.