Testing and performance of immobilized Fenton photoreactions via membranes , mats and modified copolymers

During the last 6 years our laboratory has developed Fenton immobilized catalysts for the partial or total destruction of toxic organic compounds and their mixtures. This paper reports on Fe-supported noncorrosive supported membranes and fabrics like: Nafion, Nafion-glass mats and polyethylene block copolymers. These novel supported catalysts have shown acceptable kinetic rates, resistance to the leaching of Fe3+ into the solution and no corrosion to the highly oxidative radicals generated in the solution during Fenton immobilized photo-assisted catalysis. Nafion-Fe membranes degrade Orange II under visible light only up to pH 4.8. In the case of nafion glass mats supported Fe3+-ions, the initial pH could be raised up to 8 or above. The pH decreased to about 4 during the photodegradation of Orange II due to the formation of intermediate carboxylic acids but the costly initial acidification process necessary in the case of homogeneous Fenton processes is avoided. Carboxylates and carboxylic acids were observed by IR spectroscopy on the surface of the supported catalysts towards the end of the photodegradation process as well as carboxylic acids detected by HPLC. The IR bands are found at 1523 and 1557 cm−1 in the case of the copolymer-Fe3+ corresponding to two types of iron-carboxylate species. The formation of carboxylates explains the drop of pH during the photodegradation to values between 3 and 4 corresponding to the pKa of the carboxylic functional group.


INTRODUCTION
During the last decade hydroxyl radical ( • OH) chemistry has attracted wide interest in oxidation technologies in wastewater treatment and the Fenton reagent [1] has been commonly used as source of • OH radicals in acidic media Recently [2] it has been shown that UV-Vis light accelerates reaction (1) improving the degradation rates of many organic compounds e.g.; the non-biodegradable azo-dye Orange II taken a model compound in this study [3,4].Iron salts as shown in eq. ( 1) have been known to be the most common catalysts for H 2 O 2 decomposition in acidic media.They have been used under light irradiation for this purpose in concentrations of 20-50 mg Fe-ion/L and in dark reactions at much higher concentrations.Only in this way a suitable H 2 O 2 decomposition kinetics with meaningful yields of • OH-radicals as shown in eq. ( 1) were attained in photo-assisted Fenton reactions [5,6].Fe-sludge disposal and/or regeneration after the Fenton reactions is a serious problem during pollutant degradation in homogeneous media.The removal of iron-ions is a relatively simple operation commonly carried out by precipitation and re-dissolution of the Fe-ions after the treatment of large volumes but it implies the use of large amounts of chemicals and manpower.This costly regeneration step has therefore to be avoided.After the Fenton treatment of waste-waters the remaining Fe-ions are found at concentrations far above the levels allowed in waste waters by the EEC regulations of 2 ppm [1].
In order to overcome such drawbacks we undertook the present study.In this study we sought to explore the possibility of highly dispersed Fe-ions supported on to dissolved Nafion membranes having similar effect as Fe-ions on the decomposition of H 2 O 2 in homogeneous solution.The immobilization the Fe-ions on a membrane allows the catalysis of the disappearance of Orange II under light in a novel as compared to homogeneous solutions avoiding the drawbacks of the disposal of the iron-ions at the end of the treatment.This new approach will be applied to a non-biodegradable azo-dye (Orange II) found in the effluents of the textile industry [8,9] as part of the studies conducted by our group in the emerging field of Advanced Oxidation Technologies (AOT's).
Membrane related research has attracted much attention in recent years as a structured medium for photochemical reactions.Nafion perfluorinated membranes (Dupont) have been used during the last decade in a variety of catalytical/electrocatalytical integrated chemical reactions [10].Polymer membranes have also been used for charge transfer in inorganic and biological systems [11].Very few studies of chemical transformations on photo-activation loaded membranes containing CdS or TiO 2 [12][13][14][15] reactions have been reported until now.Reactions on Fe-loaded polymer membrane systems have recently reported [16].In the present study the non-biodegradable azo-dye Orange II has been chosen as a probe and its degradation in aqueous solution has been investigated.Azo-dyes account for more than 22% of the world dye production and are commonly found in effluents of the textile industry [5][6][7].The hydroxylation/oxidation of Orange II [3,4] provides a suitable test for the oxidative degradation of this material under conditions relevant and technical systems.The pretreatment proposed in on the supported materials described in this study has shown to be effective for subsequent biological treatment-a low cost process-to complete the abatement of pollutants.

Nafion-Fe 3+ membranes as photocatalysts.
Figure 1  concentrations greater than 0.8 mM, no further acceleration was observed occurring during Orange II degradation.This is due to the known scavenging effect when using higher H 2 O 2 concentration on the further generation of hydroxyl radicals in solution [8].The inset in Figure 1 shows a more detailed kinetics at times < 200 min.The concomitant total organic carbon (TOC) decrease was observed to be modest going from 16 to 11 ppm C after 120 min.Within this time the azo-dye is observed to completely disappear from the solution.Concomitant CO 2 evolution was checked by gas chromatographic technique (GC) confirming that mineralization of the dye proceeds with a much slower kinetics than the Orange II disappearance.This reveals the formation of longer lived intermediates in solution.This corresponds to about a fourfold increase in the OH − concentration in the solution and suggests that reaction (2) as the main pathway for Orange II degradation and not reaction (3) Three experimental observations substantiate further the mechanism suggested in eq. ( 1): a) methanol (0.26 M) precluded the abatement of the azo-dye observed in Figure 1 due to its • OH radical scavenging properties, b) spectrophotometric results showed that the initial Nafion/Fe(III) membrane decolored substantially during the photolysis due to the build-up of Nafion/Fe(II) absorbing at much lower λs.Since the iron has been exchanged at room temperature in aqueous solution the exchanged-hydrated Nafion/Fe(III) would involve the presence in the Nafion of the Fe-species: Fe(OH) 2+ ε 366 nm = 275 M −1 cm −1 and Fe 2 (OH) 2 4+ ε 366 nm = 1000 M −1 cm −1 [18].During the dye decoloration (shown in Figure 2) the orange-brown color of Nafion/Fe(III) gradually disappears due to the formation of the much lower absorption from the charge transfer band of Fe(H 2 O) 6 2+ observed below λ = 265 nm with ε 254 nm = 20 M −1 cm −1 [19] and finally c) if the superoxyde radical HO 2 • (pK a = 4.8) would be photo-produced in eq. ( 2) the latter radical has a considerably lower one electron standard reduction potential of E 1).The fast oxidation of the Orange II observed in Figure 1 cannot possibly be explained in terms of this kinetically slower and less energetically HO 2 • radical (eq.( 2)).
Figure 2 shows the repetitive nature of the Orange II photodegradation in the presence of of Nafion/Feloaded membrane and H 2 O 2 .After about 25 cycles, the membrane was regenerated by immersion in a solution NaOH 1 M.The timing of this regeneration was determined by two factors: a) During use, the initial orangebrown coloration with an absorption (up to λ = 600 nm) slowly changed to gray-yellowish with a spectral absorption reaching only the upper limit of λ ∼ 410 nm.This indicated that during the photolysis the Fe(III) content in the Nafion membrane decreased leading to the formation of the colorless Fe(II)-ion since the total iron content of the membrane was seen to remain constant (see text below).The Nafion without any Feloading absorbed only below 300 nm and b) during the degradation cycles the membrane become kinetically faster with time.That is the degradation cycles in Figure 2 occur within less time and Orange II degradation in times shorter than ∼ 40% were observed compared to first degradation run.Nafion/Fe(II) highly stable species are probably responsible for the observed process acceleration as the reaction progresses in time.
The assessment of the oxidation state of the iron during the reaction was carried out by photoelectron spectroscopy (XPS) in a Leybold-Heraeus instrument.The quantitative evaluation of the Fe(III) and Fe(II) species before and after 120 min light reaction revealed 78% Fe(III) at 710.0 eV(B.E.) and 22% Fe(III)/Fe(II) species at 713.8 eV(B.E.) at time zero.The former species appeared mainly as Fe 2 O 3 while the latter species consisted mainly of the composite Fe 3 O 4 oxide showing the peaks for the constituent oxides.After reaction, the Nafion membrane showed 16% of Fe(III) and 84% of the Fe(III)/Fe(II) oxidation states.The corrections for electrostatic charging of the particles during the measurements was carried out by the polynomial fit of the data with a Shirley-type background subtraction.The XPS experimental result further confirms the spectral change in the Fe-Nafion membrane observed during the photocatalysis.The size of the Feclusters on the Nafion was determined by transmission electron microscopy (TEM) to be 37± 4 Å by way of a Philips 20 M instrument.The particles were seen to be uniformly deposited on the Nafion and the size distribution was narrowly centered around the median value The inset in Figure 2 shows the change in absorption of the Nafion membranes with time.After 500 hours irradiation a substantial change in absorption of the Nafion/Fe(III) with respect to the absorption observed at time zero.A decrease in the absorption of Nafion/Fe(III) is observed along the concomitant growth of the lower absorbing Nafion/Fe(II).This change in optical density is consistent with the molar absorption coefficients and the XPS evidence presented above in this study.The optical absorption of Nafion membranes is shown in the left-hand side of the inset.The formation of Q-sized Fe(III)/Fe(II) clusters in the Nafion upon irradiation has been confirmed by X-ray diffraction (XRD).The exchange procedure used to incorporate Fe-ions in the membrane is therefore also a method to produce nano-sized particles of iron oxides embedded in the polymer structure.
Based on the experimental results outlined above in the text to (Figures 1 and 2) a simplified scheme of reaction is presented in Scheme 1. Scheme 1 shows the photocatalytic degradation of Orange II in acidic solutions due to the photoproduction of oxidative radicals in solution and the build up of Nafion(Fe(III)/Fe(II))species during the degradation process.
More recently [21], Orange II abatement in a membrane based reactor has been reported.Nafion/Fe 3+ (1.78%) membrane was used under Blue Actinic light (366 nm and 36 W) in a 1.20 long concentric immersion reactor.The degradation up to bio-compatibility of the azo-dye Orange II proceeds at acceptable kinetic rates but only up to pH 4.7 as reported in ref. [21].The membrane was left to freely float between the light and the walls of the reactor system shown in Figure 3.The intermediates of the degradation were observed to be bio-compatible after the initial scission of the −N = N− bond of Orange II.

Nafion/ Fe 3+ glass mats as heterogeneous photocatalyst for Orange II abatement. The azo-
dye Orange II has been taken as a model azo-dye to monitor the decoloration and degradation in photo-Fenton reactions on Nafion/Fe 3+ membranes [22].These processes were possible only up to pH 4.7 since at this pH apparently active components of the Fecluster on the Nafion membrane deactivate due to precipitation [22].
More recently [23] Fe-ions were exchanged on Nafion droplets that were later attached to glass-mats.The Nafion oligomer containing the Fe-ions achieved a suitable dispersion on the glass-mats and the degradation of the model azo-dye was possible.Harmer et al. [24], has shown that acid groups of the Nafion are more accessible in Nafion-glass composites.The accessibility of these sulfonic groups leads to an improved ion-exchange capacity at the interface between Nafion and the glass-mat.Therefore, different Fe-clusters may be present on the Nafion-glass composites compared to Nafion/Fe 3+ membranes.The decoloration/degradation of Orange II proceeded at a rate close to the one observed on Nafion/Fe 3+ [22] and the catalyst could be recycled without loosing his efficiency many times showing the stable nature of the process under study as shown in Figure 4. Figure 4 presents the catalytic nature of the decoloration of Orange II under Suntest irradiation on Fe 3+ /Nafion/glass fibers in the presence of H 2 O 2 .At the end of each degradation cycle, suitable amounts of Orange II and H 2 O 2 were added to the original solution as selected for the photocatalysis.The results in Figure 4 confirm the photocatalytic nature of Orange II degradation.No measurable amounts of Fe 3+ -ions were found applying the thiocyanate test.The absence Fe 2+ -ions was analyzed by complexation with phenanthroline.
The mechanism of Orange II photocatalytic degradation on Nafion/Fe 3+ /glass mats under visible light irradiation is shown in Scheme 2.
Repetitive mineralization of Orange II under Suntest irradiation on Fe 3+ /Nafion was observed to proceed up to pH 4.7 [22].But when Fe 3+ /Nafion/glass mats were taken as photocatalysts the decoloration of Orange II solutions was observed to occur in the pH range 2.8-9.0 as shown in Figure 5. Since with Fe 3+ /Nafion/glass mats degradation of Orange II is possible up to pH 8 as shown in Figure 5, we should save the cost of acidification of effluents with pH 6-9 when applying photo-Fenton treatment and only would have to add base at the end of the pre-treatment to attain again pH 6 to couple the Fenton pretreatment with biological secondary processing.Acidification is costly and exceeds the cost of the oxidant and the energy used during the Fenton treatment of large volumes of waste waters.
Figure 6 shows decrease in the total organic carbon content (TOC) for Orange II solutions having different dye concentrations.The mineralization in the dark is seen to be < 10% for solutions Orange II (0.20 mM) due to the decomposition of H 2 O 2 by the Fe 3+ /Nafion/glass mats.Addition of H 2 O 2 to Orange II (0.20 mM) did not lead to any observable mineralization in the absence of Fe 3+ /Nafion/glass mats.The mineralization of Orange II is seen to proceed under Suntest light irradiation for different concentrations of Orange II in Figure 6.The steeper decline in the TOC observed at higher Orange II concentrations is due to the more favorable mass transfer taking place between the solution Orange II and the Fe 3+ /Nafion/glass fibers during reaction.The mass transfer between the solution and the Fe 3+ / Nafion/glass fibers is driven by the difference in Orange II concentration existing between the bulk of the solution and the glass mat interface.The diffusion length (x) of the oxidative-radical away from the glass mat can be estimated from the Smoluchowski diffusion relation The reaction rate between the OH • radical and Orange II has been recently reported to be close to 10 9 M −1 s −1 [25].At a concentration of Orange II (0.20 mM), the lifetime of the encounter pair is ∼10 −6 s.With D ∼ 5×10 −6 cm 2 /s, a value for the diffusion length (x) of ∼ 50 nm eq. ( 4) is found for the OH • radical.In the case of the HO 2 • radical, the value of the reaction rate of HO 2 • with Orange II is found to be ∼ 10 6 M −1 s −1 .From this a diffusion length of ∼ 310 nm can be estimated for the HO 2 • radical away from the Fe 3+ /Nafion/glass.Orange II.Oxalic, formic and acetic and smaller concentrations of other acids have been reported during the mineralization of Orange II [26].The pH of the solution at the end of the degradation was observed to be in the region to the pK a of the short carboxylic acids already mentioned.More interesting the final pH of the solution after Orange II degradation on Fe 3+ /Nafion/glass mats attains the value corresponding to the pK a of the carboxylate group independently of the initial pH of the Orange II solution.This is shown in Figure 7.This is interesting since Fenton reagent Fe 3+ /H 2 O 2 does not produce by itself a decrease of the pH in the aqueous solution when generating the OH • and HO 2 • radicals leading to the destruction of the organic compound as shown in Scheme 3. Therefore, the degradation of Orange II involves the formation of Fe-organic complexes is noted as where the LMCT complex in eq. ( 5) is the precursor step leading to the abatement of Orange II.Equation ( 5) proceeds concomitantly to the radical reaction contributing to the organic compound degradation Fe-organic complexes or Fe-chelates intermediates like oxalates, maleates, pyruvates mentioned above in eq. ( 5) and decompose to CO 2 with accelerated kinetics in Fenton photo-assisted reactions.But this decomposition would not occur due to the slow kinetics in dark reactions as observed during Fe-immobilized Fenton reactions mediating the photo-decomposition of Orange II [26] and coumaric acid [27].Scheme 4 shows the competitive reactions taking place under light and in the dark.Nothing is practically known about the nature of the detailed nature of Fe-carboxylates in the presence of H 2 O 2 although some studies have reported the existence of hydroperoxides [28] and unstable peroxyradicals [29] during Fenton traditional catalytic oxidation processes.
The pH of the Orange solution drops from 6 to ∼ 4 during mineralization

Modified block-copolymers thin films loaded
with Fe 3+ and TiO 2 .More recently [30], Fe 3+ , Fe 2 O 3 and TiO 2 have been immobilized on low cost polyethylene modified copolymers films containing maleic anhydride anchoring groups.The observed rates of degradation of Orange II and halocarbons were only slightly below the rates observed during homogeneous Fenton photo-assisted degradation when Fe 3+ was exchanged on the copolymer surface.Polyethylene is known to be the second most inert Dupont polymer after Teflon, a C − F polymer of widespread use.IR spectroscopy suggests that Fe 3+ /Fe 2 O 3 interacts with the negatively charged conjugated carboxylic groups through −COO − Fe 3+ .In the case of the copolymer-Fe 3+ , the bands in the IR spectrum at 1523 and ∼ 1557 cm −1 correspond to two types of iron-carboxyl species.The strong attraction between the negative anhydride groups on the polymer surface and the Fe 3+ -ion leads to formation of metal to ligand transfer charge bond (MLTC) between the negatively anhydride group and the Fe-ion on the copolymer thin films.This accounts for the lack of Fe 3+ ions leaching into the solution.It is also the reason for the attachment of a monolayer of complexed species on the thin film surface.During the last decade it has been repeatedly invoked that homogenous Fenton processes involve three reactive intermediate species: • and other oxidative radicals available in solution.But this was not the case for [polyethylene-COO − − Fe 3+ ] even when used over long times (300 hours) and/or repeated recycling below the polyethylene flowing temperature (80 • C).Moreover, the observation reported recently [31] that degradation of organic compounds is able to take place during Fenton photo-assisted treatment in the presence of 3000 ppm of Cl − -ion involving also the formation of Cl 2 −• radicals lends further support to the existence of Fe-carboxylate complexes in solution during the mineralization of organic compounds to CO 2 .

CONCLUSIONS
The results presented in this study show that supported catalysts are a practical alternative to homogeneous systems or suspended catalytic devises in the field of pollution abatement.The supported catalyst should have three inherent characteristics to be viable: a) not leach out the catalytic metal-ion or semiconductor, b) not be corroded by the oxidative radicals generated in solution and c) intervene in the degradation of the pollutant(s) with an adequate kinetics.The last condition is the most difficult to meet during the treatment of real industrial wastewaters.Our research program is currently designed to improve the kinetics of this type of processes.

Figure 6 .
Figure 6.Mineralization of Orange II solutions of different concentrations on Fe 3+ /Nafion/glass fibers in the dark and under Suntest light irradiation (80 mW/cm 2 ) in the presence of H 2 O 2 (10 mM) at pH 3. Open points denote runs under light and the full points refer to the run in the dark.

AFigure 7 .
Figure 7. Variation of pH with time on Nafion/ Fe 3+ /glass mats for Orange II and other organic compounds under light irradiation.

2 •
a) the formation of oxidative radicals like OH • , HO and other oxidative radicals due to the H 2 O 2 added in solution, b) the possible formation but controversial solution species Fe(IV) and finally c) the formation of Fe-chelates as shown in Scheme 4.The importance of the [R − COO − ] as a preferred reaction pathway in the decomposition of pollutants on the surface [polyethylene-COO − − Fe 3+ ] is rationalized by the observation that the copolymer thin films should be attacked by OH • , HO 2 shows the disappearance of Orange II (pH 2.8) under light irradiation on a Fe-loaded Nafion membrane COOH] . OH, HOO .,O2• OH, HOO • , O 2 • OH, HOO • , O 2