Vitória M.Almeida,Carla A.Orge,M.Fernando R.Pereira,O.Salomé G.P.Soares

Laboratory of Separation and Reaction Engineering -Laboratory of Catalysis and Materials,(LSRE-LCM),Faculdade de Engenharia da Universidade do Porto,Rua Dr.Roberto Frias,4200-465 Porto,Portugal

Keywords:Advanced oxidation processes Catalysts Emerging pollutants Ibuprofen

ABSTRACT The degradation of the anti-inflammatory ibuprofen (IBP) was evaluated by several advanced oxidation processes.IBP was treated by single ozonation and oxidation with hydrogen peroxide (H2O2),as well as a combination of these treatments.In order to improve the efficiency,the presence of catalysts such as original carbon nanotubes,labelled as CNT,and iron oxide supported on carbon nanotubes,named as Fe/CNT sample,was considered.The evolution of IBP degradation,mineralization and toxicity of the solutions was assessed.The formation of intermediates was also monitored.In the non-catalytic processes,IBP was faster removed by single ozonation,whereas no significant total organic carbon (TOC)removal was achieved.Oxidation with H2O2 did not present satisfactory results.When ozone and H2O2 were combined,a higher mineralization was attained(70%after 180 min of reaction).On the other hand,in the catalytic processes,this combined process allowed the fastest IBP degradation.In terms of mineralization degree,the presence of Fe/CNT increases the removal rate in the first hour of reaction,achieving a TOC removal of 85%.Four compounds were detected as by-products.All treated solutions presented lower toxicity than the initial solution,suggesting that the released intermediates during applied processes are less toxic.

1.Introduction

Water is a natural and indispensable resource for the existence of life on earth,and it is necessary to safeguard its quality and reduce the levels of contamination.Contaminant compounds are released every day to water resources.In the last years,the occurrence of compounds,in small concentrations,has been labelled as emerging pollutants.Pharmaceutical compounds,like ibuprofen(IBP),are an example of this class of pollutants.IBP is a nonsteroidal anti-inflammatory substance derived from propionic acid and a broad spectrum of action,making it one of the most consumed drugs worldwide[1].IBP has a strong presence in the aquatic environment due to its complex degradability,which leads to the formation of several intermediates that are not able to be completely removed by conventional treatments,such as 2-hydroxyibuprofen [2].In Portugal,the consumption of IBP has been increasing and its detection in aquatic ecosystems.IBP was detected in several Portuguese rivers in concentrations in the hundreds of nanograms per liter.The main points of contamination identified were hospitals,WWTPs,and landfills,the latter being the major contributor to the release of IBP to aquatic matrices[3].Therefore,it is necessary to apply alternative treatments,such as advanced oxidation processes (AOPs),to remove effectively IBP from the waters.

AOPs are oxidative methods based on the formation of highly reactive and oxidizing species such as hydroxyl radicals (),which rapidly oxidize several organic compounds.Its oxidation potential(E0=2.80 V)is higher than that of other conventional oxidants,thus becoming an extremely efficient species in the oxidation of a wide range of organic compounds [4].The great advantage of AOPs is that they can achieve complete mineralization,which means that compounds and their intermediates are completely oxidized into carbon dioxide,water,and other simple inorganic compounds.However,in some cases,these processes can only convert the pollutant by-products,more easily degradable,and sometimes less harmful to water resources [5].Ozone(O3)is widely used as a powerful oxidizing agent(E0=2.07 V)capable of rapidly reacting with a wide range of organic compounds,but in most cases,complete mineralization was not achieved [6].The O3molecule can oxidize pollutants either by direct and selective mechanism or by an indirect mechanism involving chain reactions due to O3decomposition until the production of OH radicals[7].O3tends to selectively react with unsaturated compounds(alkenes,alkynes,aromatic rings)[8],but complete mineralization does not usually occur,remaining in solution aldehydes,ketones,alcohols,and carboxylic acids[9].The occurrence of direct or indirect reactions is mainly affected by the pH of the solutions.Indirect reactions,which involve the formation ofradicals,occur preferentially at alkaline pH,while direct reactions through molecular O3occur mainly at acid pH [10].Hydrogen peroxide (H2O2) is another oxidizing agent widely used in AOPs.H2O2may act as an oxidizing agent,but its oxidation power(E0=1.78 V)is lower than O3orradicals.For an increase in the efficacy of H2O2,it can be combined with O3,radiation and/or metal salts.It is expected that these combinations will lead to a higher generation ofradicals,which has a high oxidizing power [11].Therefore,combination processes can reduce operating cost,reaction time to achieve the same degradation efficiency with an increased rate of degradation and mineralization of recalcitrant organic compounds present in the wastewater [12].

Recently,some researches about the degradation of IBP by different AOPS were reported.Al-Hamadaniet al.[13] studied the degradation of several pharmaceuticals,including IBP,using an ultrasonic reactor in the absence and presence of single-walled carbon nanotubes.As expected,the removal of IBP was enhanced when the carbon material was added by improving oxidation species and adsorption on nanotubes.Sonochemical based processes were tested for the removal of IBP in different water matrixes and under various operating parameters[14].Coupling ultrasound with Fenton reaction produced a positive synergy effect,especially in terms of mineralization yield,whereas no significant beneficial effect was observed adding only H2O2.Additional experiments revealed that the regeneration of ferrous ions is enhanced by ultrasound,which can justify the observed results.Degradation of IBP by heterogeneous Fenton oxidation using Fe/ZSM5 catalyst was investigated under different operational conditions [15].Under best conditions,88%of IBP was removed after 3 h of reaction under natural pH conditions,but no significant mineralization was achieved (only 27% of TOC (total organic carbon content)).Very low leaching and negligible activity of leached iron in solution were observed,indicating that the Fenton reaction was mainly induced by iron species on the catalyst surface.Saeidet al.[16]studied the removal of IBP by non-catalytic and catalytic ozonation runs in a semi-batch reactor.About 93% of IBP was degraded after 4 h under optimal conditions.The addition of tested zeolites produced a significant improvement in the degradation rate,probably due to the amount of Br?nsted and Lewis acid sites.

In this work,O3based processes were assessed for IBP degradation.The efficiency of different treatments was detailed studied analyzing several parameters as IBP concentration,mineralization level,intermediates formation (hydroquinone -HDQ;1,4-benzoquinone -BZQ;oxalic acid -OXL;pyruvic acid -PRV) and toxicity of released by-products.According to the literature,it is clear that the use of O3,alone or combined with other oxidants,is efficient for IBP degradation[17,18].However,most of the studies only evaluate the concentration of IBP during the processes,and most of them only reported homogeneous treatments.In this work,a more detailed study was reported.In addition to IBP concentration evaluation,several parameters during different treatments were analysed in order to investigate the efficiency of the processes in terms of intermediates released and respective toxicity.The effect of the addition of heterogeneous catalysts to selected oxidants was considered to improve the mineralization level of the treated solutions.Accordingly,combinations of O3with H2O2and/or catalysts were considered,and two catalysts were selected:carbon nanotubes,labelled as CNT,were commercially obtained,and iron supported on carbon nanotubes,named as Fe/CNT.The catalytic activity of two selected samples during individual and combined processes was assessed in terms of pollutant removal,mineralization,released by-products and toxicity of treated solution.To perform the comparison between the processes,the initial concentration selected was higher than the values usually present in real conditions.The catalysts were chosen according to their activities for tested processes.Subsequently,all catalysts were characterized by different techniques in order to analyse their properties.Carbon nanotubes were selected due to their strong catalytic activity with O3,especially for promoting the O3decomposition and surface reactions.CNT is a mesoporous material,consisting of an agglomeration of tubes,avoiding mass transfer limitations that are usually found in microporous materials [19].On the other hand,the decomposition of H2O2into radicals is promoted in the presence of Fe,as described in Eq.(1).

H2O2can also participate in the formation of other radicals with O3,as described in Eq.(2) [11].

In the case of Fe/CNT sample,the presence of CNT avoids hydrolysis and iron precipitation,while iron promotes the increase of the number of active centers on the catalyst surface;thus,the formation ofradicals increases and,consequently,the oxidation of intermediates formed during the reaction is promoted,leading to higher mineralization [20,21].Additionally,O3adsorption and reactions on the surface of the CNT can occur,and generation of surface oxygen radicals is promoted,justifying the importance of the textural properties in these processes [22-24].

The presence of synergetic effects was evaluated when O3and H2O2were combined in solution,both in the absence and presence of tested catalysts.The catalyst stability was studied.

2.Material and Methods

2.1.Preparation and characterization of catalysts

Multi-walled carbon nanotubes (CNT) were used as received from the Nanocyl company.Fe/CNT sample was prepared by incipient impregnation method.In this procedure,a solution of Fe(NO3)3·9H2O was added dropwise through a peristaltic pump to obtain a mass content of 2% iron under ultrasonic mixing.After impregnation,the material was dried for 24 h at 100 °C.Finally,the solid was thermally treated,firstly under nitrogen flow(400°C,1 h),and secondly under hydrogen flow(400°C,3 h).This temperature was previously determined by the temperatureprogrammed reduction(TPR) technique.

The selected catalysts were characterized by nitrogen adsorption isotherms at-196°C in a Quantachrome NOVA 4200e apparatus.Fe/CNT catalyst was also characterized by TPR,whose analysis was held in an AMI-200 (Altamira Instruments).Images of transmission electron microscopy(TEM)were collected in a LEO microscope model 906E with an acceleration voltage of 120 kV.The pH at the point of zero charge (pHPZC) was determined by mixing 0.02 g of the catalysts with 20 ml of NaCl solution (0.01 mol·L-1).The pH was adjusted with HCl or NaOH solutions (0.01 mol·L-1),in order to obtained values between 2 and 12.The final pH was measured after 24 h under stirring at room temperature.The pHPZCvalue of each carbon sample was determined when the curve pHfinalvs.pHinitialcrosses the line pHfinal=pHinitial.

2.2.Kinetics experiments

Experiments of single ozonation(O3)and oxidation with hydrogen peroxide(H2O2),as well as the combination of these processes(O3+H2O2) were evaluated in the degradation of ibuprofen (IBP),in the absence and presence of a catalyst,and all reactions were carried out in the same glass reactor with a capacity of 250 ml.The experiments were carried out under optimal conditions previously determined.The IBP initial concentration used was 20 mg·L-1.This concentration was selected according to the analytical methods available,and in spite of this concentration is higher than verified in real conditions,it allows to make a better comparison between the different processes,and different catalysts,and to study the mineralisation degree by TOC measurements.Thus,the evaluation of the mineralization level achieved under high concentration is important to evaluate if the intermediates were also removed.The tests were performed at the natural pH of the initial solution (≈6.2),in order to avoid adding extra chemicals for pH control.The pH of initial and treated solutions was measured in pH meter WTW series from InoLab.The agitation was maintained constant at 400 r·min-1.In O3based reactions,O3was generated from pure oxygen through the BMT 802X generator.The concentration of O3was monitored by a BMT 960 analyzer.The O3flow rate and concentration used were 150 cm3·min-1and 50 g·m-3,respectively.The gas was introduced in the reactor by a diffuser (d=1 cm).O3that non-react in the gas phase left the reactor by a gas washing bottle filled with potassium iodide solution to abate ozone by oxidation of iodide to iodine.In the reactions without O3,the gas used was O2,in order to guarantee the same experimental conditions.In the oxidation reactions with H2O2(30%(vol)),123 μl,calculated based on equation 1,was added using a 50% excess of H2O2.A certain amount of sodium sulfite(Na2SO3) was added to collected samples,according to Eq.(2),in order to guarantee the total consumption of peroxide,thus stopping the reaction.

In the catalytic reactions,125 mg of catalyst was added.The collected samples were centrifuged for 15 min with a speed of 13,500 r·min-1in the centrifuge (VWR MicroStar 12),in order to guarantee the absence of suspended particles.

The amounts of Fe eventually leached during the reaction were measured in a UNICAM 939/959 atomic absorption spectrometer(AAS)(detection limit-0.18 mg·L-1),using the remaining solution after tested processes in the presence of Fe/CNT sample.

The replicated tests demonstrated that the process is reproducible,and no differences were observed.

2.3.Analytical methods

The evolution of the concentration of IBP and intermediate compounds was evaluated by high-performance liquid chromatography (HPLC).Detection of IBP,HDQ and BZQ was performed on a Hitachi Elite Lachrom HPLC equipped with a diode detector(DAD).For that purpose,a LiChroCART Purospher STAR RP-18(250 mm × 4.6 mm) column with a mobile phase consisting of sodium dihydrogenphosphate solution (pH=2.8) and methanol(30/70)were used at a flow rate of 1 ml·min-1and an injection volume of 50 μl.The concentration of carboxylic acids,OXL and PRV,was determined on a Hitachi Elite LaChrom HPLC equipped with UV detector,using an Alltech AO-1000 column (300 mm ×6.5 mm)with a mobile phase of 5 mmol·L-1H2SO4solution,a flow rate of 0.5 ml·min-1and an injection volume of 15 μl.The degree of mineralization was evaluated by measuring the total organic carbon content (TOC).The analyzes were performed using the NPOC method in Shimadzu’s Total Organic Carbon Analyzer TOC-L,coupled with the ASI-L autosampler of the same brand.The toxicity of the treated solutions was evaluated using Microtox?acute toxicity tests.This technique is based on the evaluation of the inhibition of the luminescence of the Gram-negative marine bacteriumVibrio fischeriwhen it is submitted to the presence of test substances for 30 min at 5 °C.

3.Results and Discussion

3.1.Catalysts properties

More information about catalysts characterization can be found in our previous work [19].The textural properties of tested materials were obtained from nitrogen adsorption isotherms at-196 °C.Sample CNT presents a surface area of 197 m2·g-1and the presence of iron does not change the surface area(196 m2·g-1) once a low amount of Fe was introduced(2%(mass)).The catalysts present a basic character with a pHpzcof 7.

TEM analysis (Fig.1) confirmed the presence of iron on CNT,showing particles with uniform size and homogenous distribution.

With the aim of identifying the reduction temperature of the metal,a TPR analysis of Fe/CNT was performed under 5%of hydrogen flow (results not shown).Two reduction peaks were detected,the first peak between 300 and 400 °C,corresponding to Fe3O4metal reduction,and the second peak between 400 and 600 °C,where the metal was sequentially reduced to FeO and Fe0[25].The real amount of Fe on support was calculated by TGA(thermogravimetric) analysis,and 2% of metal supported on CNT was obtained,as expected.Although the samples had been reduced under hydrogen at 400 °C,iron was found only in the form of Fe2+and Fe3+by XPS (X-ray photoelectron spectroscopy).

3.2.Non-catalytic processes

The evolution of IBP degradation and TOC removal during noncatalytic processes are presented in Fig.2.

Single ozonation enabled rapid removal of IBP,showing a complete depletion after 15 min of reaction.IBP is constituted by an aromatic ring,double bonds,and delocalized electrons that increase the reactivity with O3,thus becoming a perfect molecule for the direct attack of molecular O3.Quero-Pastoret al.[1]observed a similar degradation profile,obtaining a degradation of 99% after 20 min of reaction.On the other hand,single ozonation did not show complete mineralization,leading to approximately 44% TOC removal after 3 h of reaction.This result can be justified by the presence of compounds resistant to O3attacks,such as carboxylic acids.The molecular O3can not completely oxidized saturated short-chain compounds.Thus,they remain in solution [26].This result is expected and is in accordance with previous works[27,28].Molecules as metolachlor,a widely used herbicide detected in surface water and groundwater in μg·L-1,and sulfamethoxazole,a synthetic antibiotic also found in water bodies in small concentrations,are easily removed by O3alone.Still,the mineralization achieved by single ozonation was not satisfactory.

The oxidation with H2O2did not lead to the degradation of IBP,with no decrease in concentration over time and,as expected,TOC analyzes confirmed the process inefficiency for IBP degradation,since ratio TOC/TOC0remained constant and close to 1 during all the reaction time (3 h).H2O2has lower oxidant power than the others (,O2,and O3) and its direct oxidation reaction with organic pollutants is generally less inefficient because its radical decomposition is slow[26].Treatment with H2O2can be improved with increasing temperature since H2O2decomposition is promoted,as well as the increase in the concentration used.

Fig.1.TEM images of (a) CNT and (b) Fe/CNT samples.

Fig.2.Evolution of dimensionless IBP (a) and TOC (b) concentration during non-catalytic processes.

The combination of O3with H2O2did not show any benefits to the single ozonation during IBP degradation under tested experimental conditions.In the initial stage,O3+H2O2presented slower IBP removal than O3alone;however,a total removal was achieved after 15 min of reaction.This result is not completely unexpected,since in the case of compounds with high reactivity to ozone attack,as IBP,the improvement ofradicals in solution does not always increase the pollutant degradation rate once less ozone is available by the enhancement of decomposition of ozone intoradicals [28].In the case of TOC removal,the combination of O3with H2O2showed higher mineralization than single ozonation,obtaining a TOC removal of 70% after 180 min of reaction.This increase in TOC removal is probably due to the more significant degradation of the intermediate compounds,since in addition to the direct O3attack,there is the formation ofradicals.Higher mineralization in the combined processes,compared to single ozonation,was also verified in other study [29].This increase is justified by the additional formation ofradicals from the cleavage of the H2O2molecules by the action of the O3molecule.

3.3.Catalytic ozonation and catalytic oxidation with H2O2

In the case of IBP removal rate by catalytic ozonation,Fig.3 a)and b),CNT and Fe/CNT samples did not present significant differences (experimental error ＜±2%),showing a complete removal after 30 min of reaction,as verified in the non-catalytic run.

In terms of mineralization degree,the addition of tested catalysts to O3had a positive effect,leading to approximately 50% of TOC depletion after 3 h of reaction,in contrast with 40% achieved by the non-catalytic run.These results suggested that the presence of Fe was not advantageous in the case of catalytic ozonation of IBP,in contrast with CNT that confirmed the high activity of the carbon phase for ozonation reactions [30].

In order to verify the influence of adsorption during catalytic ozonation,adsorption on CNT (O2+CNT) was carried out,and the results are also shown in Fig.3 a)and b).It is possible to verify that 60% of IBP was adsorbed on the CNT surface in the first minutes of the reaction.In terms of TOC removal,approximately 50%of organic matter disappeared from the solution after 3 h.The adsorption on CNT was not negligible;thus,it can be accepted as one of the reaction mechanism steps.Similar results were verified during adsorption on Fe/CNT sample due to the low amount of Fe introduced on CNT.According to Janset al.[31],catalytic ozonation in the presence of a carbon material significantly promotes the initiation of the chain reactions that transform O3into more reactive secondary oxidants,such asradicals.This fact may explain the activity observed in catalytic ozonation in the presence of CNT.The presence of CNT can also promote the occurrence of adsorption and surface reactions of the O3molecules in the carbon material,leading to the formation of surface oxygen radicals [30].Another important factor in explaining the high efficacy observed in catalytic ozonation is the textural properties of catalysts,since they favor the decomposition of O3inradicals on the surface of the catalyst,due to their mesoporous nature,promoting the mineralization of the compounds.

Fig.3.Evolution of dimensionless IBP (a,c) and TOC (b,d) concentration during catalytic ozonation and oxidation with H2O2.

IBP degradation results by catalytic oxidation with H2O2are shown in Fig.3 c) and d).The combination of H2O2with the selected catalysts did not completely remove IBP,but a significant increase was verified compared to the non-catalytic run.The combination of H2O2with CNT and Fe/CNT allowed an IBP removal of 82% and 88%,respectively,after 3 h of reaction.A higher initial IBP removal rate was verified with both catalysts in the first minutes of reaction;however,after this time,the improvement observed is very low.The decrease in IBP concentration during H2O2/CNT is only slightly higher than observed in single adsorption,suggesting that the adsorption step mainly contributes to IBP removal.The degradation of IBP through the Fenton process was also studied and similar results were observed[32].Oxidation with H2O2did not remove the IBP in solution,but the subsequent addition of Fe (II) led to more effective removal because of the reaction between catalyst and H2O2,producing moreradicals in solution (Eq.(3)) that attack the organic molecules.

However,Loaiza-Ambuludiet al.[33]found that the concentrations of H2O2and ferric iron were key factors in the TOC removal during IBP degradation.In this work,we were able to verify that Fe/CNT also presented better performance than CNT in terms of TOC removal,leading to a mineralization level of 64% and 45%,respectively,after 3 h of reaction.This result confirmed that the presence of Fe had a positive effect on H2O2oxidation under tested conditions,as expected.This improvement is in accordance with our recent publication [19],where Fe-based catalyst presented a higher TOC removal rate during sulfamethoxazole degradation.

No leaching of Fe was detected(within the experimental error)at the end of the reactions using Fe/CNT sample,revealing that the catalyst is stable.

3.4.Combined processes

As shown in Fig.4,the addition of catalysts to the O3and H2O2combination increased the initial rate of IBP removal.Nevertheless,independently of the process,all IBP was removed after 15 min.When compared to the single ozonation,it was observed that the addition of H2O2presents synergistic activity with O3in the presence of catalysts.The results of TOC removal reveal that the addition of the tested catalysts improves the mineralization of the solution,reaching 81% and 85% in the presence of CNT and Fe/CNT,respectively,after 3 h of reaction.

It is expected that in the combined process catalyzed by Fe/CNT,O3can be adsorbed and decomposed on the CNT surface under Fe(II)action,whereas H2O2can give increase to radical species by the action of the catalyst and O3itself,such as theradicals.Therefore,the addition of H2O2to the catalytic process is advantageous under the experimental conditions used.However,the H2O2added to the solution can adsorb on the surface of the catalyst and modify it,reducing the available active centers,and consequently,difficult the surface reactions,as well as the formation of surface radicals,which did not happen in this work.

In terms of IBP removal,no synergetic effect was verified when O3,H2O2and tested catalysts were combined,since O3alone is enough to achieve a complete degradation in the first minutes of reaction.On the other hand,TOC removal is strongly improved when O3,H2O2and tested catalysts were carried out together in a solution.The addition of H2O2to O3in the presence of CNT and Fe/CNT achieved a better mineralization level,possibly due to the formation of a higher amount of oxidizing radicals.H2O2can be decomposed in the presence of Fe and/or CNT,as well as can participate in the formation of other radicals with O3.This enhancement is a little more evident in the presence of Fe/CNT sample.Furthermore,in Fe/CNT sample,the presence of iron increases the available active centers in the CNT,which favours the decomposition of O3in radicals and the reactions on the CNT surface.These radical species,due to low selectivity,attack a wide range of pollutants,removing a high content of organic matter presented in the solution.

3.5.Intermediates evaluation

Fig.4.Evolution of dimensionless IBP (a) and TOC (b) concentration in catalytic ozonation with H2O2.

Hydroquinone (HDQ),1,4-benzoquinone (BZQ),oxalic acid(OXL) and pyruvic acid (PRV) were identified as intermediates.Fig.5 shows the evolution of the concentration of these intermediates.As expected,during catalytic oxidation with H2O2the formation of by-products is practically nonexistent due to low activity verified in this treatment.In general,the maximum concentration of HDQ is confirmed at the beginning of the reaction,and after that the concentration decreased and remained approximately 1 μg·L-1until 180 min of reaction.The non-catalytic combined process,O3+H2O2,led to the formation of a higher amount of HDQ and after 30 min of reaction the concentration remained around 4 μg·L-1.

In the case of BZQ,it was also released at the beginning of the reaction,similar to HDQ,but in lower concentration,and at the end of the reaction it disappeared,except for the combined catalytic process.Single ozonation led to the formation of a higher amount of BZQ,but in the end it was not detected.The carboxylic acids identified presented a similar profile during IBP degradation.In the non-catalytic methods,OXL (detection limit -DL=0.5 mg·L-1) and PRV (DL=0.05 mg·L-1) were released in higher amounts than in the catalytic processes.This result is in accordance with expected,since short-chain carboxylic acids are highly refractory,and the presence of a catalyst is required to remove these compounds [34,35].In the case of single ozonation,OXL and PRV concentrations increased along with the reaction,suggesting that these acids were continuously formed but not degraded.On the other hand,a different profile was verified when O3was combined with H2O2,showing a decrease in the acids concentrations which indicates that some OXL and PRV were removed.The removal of theses acids occurred by oxidation during combined treatment since the contribution of adsorption on tested catalysts were scarcely,and can,therefore,be neglected [30].

Fig.5.Evolution of HDQ (a),BZQ (b),OXL (c) and PRV (d) concentrations during non-catalytic and catalytic processes.

In order to verify the evolution of the pH during applied treatments,the pH of the final solutions was measured.During H2O2no changes in the pH were observed since no IBP was removed during this process.On the other hand,the pH of treated solutions by single ozonation and individual catalytic processes was between 3.5 and 4,which confirmed the presence of by-products of acidic nature,such as short-chain carboxylic acids.In the combined treatments,the pH was around 5.5 (pH of distilled water),confirming the high mineralization level.These results are in accordance with the results of mineralization and the evolution of intermediates.However,it is important to mention that some intermediates were not identified and/or detected.

The evaluation of the toxicity of the compounds generated in the degradation of IBP was carried out by Microtox?acute toxicity tests,and the results obtained are presented in Fig.6.The results demonstrate that all treated solutions showed lower toxicity than the initial solution,indicating that tested systems led to the formation of less toxic intermediates than the original compound.Similar results were verified during the degradation of sulfamethoxazole by different AOPs [19].No significant effect was confirmed when adding H2O2to single ozonation in terms of toxicity.However,when tested catalysts were added,considerable improvement was observed.The solution treated with O3,H2O2,and CNT had the lowest amount of toxic compounds.Although Fe/CNT presented a higher TOC removal rate,the final solution treated with neat CNT has lower toxicity.These results suggested that the reaction mechanism depends on the catalyst and the presence of a small amount of Fe leads to the formation of different intermediates with higher toxicity than by-products released in the presence of only CNT.

Fig.6.Percentage of inhibition effect of treated solutions after 3 h of reaction.

3.6.Stability tests

Reusability experiments were carried out to evaluate the eventual deactivation during the mineralization of IBP.For that,three consecutive runs of catalytic ozonation in the presence of CNT were performed.In these experiments,after each run,the solution was filtered,CNT was recovered,washed with distilled water,and dried overnight.Reused CNT were added to a fresh pollutant solution.The results of IBP removal at 15 min of reaction and TOC removal at the end of the reaction during cyclic experiments are presented in Fig.7.

In terms of IBP removal,approximately 100% of the pollutant was removed after 15 min of reaction even after three consecutive runs.On the other hand,the amount of organic matter removal decreased mainly from the first to the second run and no significant reduction was verified from the second to the third run.In the first cycle,a TOC depletion of 52% was attained,while in the second and third cycles,37% and 31% of organic matter removal were achieved.These results suggested that during catalytic ozonation the surface of CNT was oxidized by the attack of O3molecules and other oxidative species presented in solution.However,this effect was attenuated along with the runs.The reported results are in accordance with the literature[30,36],since the most evident decrease in the performance is observed after the first use of the material,tending to stabilize for the catalysts successively reused,which suggests that the oxidation of the catalyst surface tends to saturation after successive reuses.According to our previous works,the loss of activity during reusability experiments is related to the presence of acidic groups,such as carboxylic acids,on the catalyst’s surface [30,36].Performing analysis of temperature programmed desorption (TPD) to fresh and reused CNT samples,it is possible to confirm an increase in the intensity of the released peaks[37].The catalyst tends to lose efficiency in the catalytic processes when its surface is oxidized due to successive reuse of the catalyst.The introduction of oxygenated electronwithdrawing groups is responsible for reducing the electron density on the carbon surface,leading to a decrease of the catalytic activity of the material for ozone decomposition and the ability to form hydroxyl radicals.

4.Conclusions

Fig.7.IBP and TOC removal during stability tests.

Different AOPs,based on the presence of O3and H2O2,were studied in the degradation of IBP.Although single ozonation is enough to achieve a complete IBP removal,a slow organic matter removal was verified.Therefore,the combination of O3with catalysts and other processes is required to attain high mineralization levels.Catalytic ozonation in the presence of selected catalysts achieved a TOC removal of 50% after 180 min of reaction,while the combination of O3with H2O2allowed a removal of 70%.An improvement was verified when tested catalysts were added to the combined process.The faster TOC removal is confirmed in the presence of Fe.

The intermediates released during the combined catalytic process disappeared at the end of the reaction.The toxicity of all treated solution is lower than the initial solution,suggesting that tested techniques formed intermediates with lower toxicity than the parent pollutant.The lowest amount of toxic compounds was verified in the treated solution by O3,H2O2,and CNT.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was financially supported by Base Funding -UIDB/50020/2020 of the Associate Laboratory LSRE-LCM -funded by national funds through FCT/MCTES (PIDDAC) and Project PTDC/EAM-AMB/31337/2017 -POCI-01-0145-FEDER-031337 -funded by FEDER funds through COMPETE2020-Programa Operacional Competitividade e Internacionaliza??o (POCI) and with financial support of FCT/MCTES through national funds (PIDDAC)and by NORTE-01-0247-FEDER-069836,co-funded by the European Regional Development Fund(ERDF),through the North Portugal Regional Operational Programme (NORTE2020),under the PORTUGAL 2020 Partnership Agreement.CAO acknowledges FCT funding under DL57/2016 Transitory Norm Programme.OSGPS acknowledges FCT funding under the Scientific Employment Stimulus -Institutional Call CEECINST/00049/2018.