Efficiency and toxicity: comparison between the Fenton and electrochemical processes

The textile industry is an industrial sector of great socioeconomic importance in terms of production and employability. On the other hand, the various processing stages involved in textile manufacture consume a large amount of water, dyes, and other chemical compounds, generating effluents that cause serious environmental issues when discharged without adequate treatment (Dasgupta et al. 2015). The strongest environmental impact of the textile sector is related to primary water consumption (80–100 m³ ton⁻¹ of finished textile) and wastewater discharge (115–175 kg of chemical oxygen demand (COD) ton⁻¹ of finished textile). This sector releases a wide range of poorly biodegradable organic chemicals into water sources, thereby affecting water salinity and color.

Among the pollutants present in textile industry effluents, dyes are certainly the main source of contamination: part of these pollutants (from 2 to 50%) end up in wastewater after the dyeing process. Dyes consist of highly complex and poorly biodegradable molecules. Their discharge into water bodies gives rise to aesthetic issues and hinders light penetration. Moreover, the carcinogenic, toxic, and mutagenic effects of dyes and their corresponding sub-products cause problems in aquatic organisms (Punzi et al. 2015). Water resources eutrophication is the ultimate consequence of dye disposal into the environment.

The high quantities of salt and fixing agents used during textile dyeing and weaving make textile industry effluents even more complex. Therefore, finding more efficient treatment methods to degrade these contaminants and their sub-products in effluents is an urgent matter (Gupta et al. 2015; Punzi et al. 2015).

The scientific literature contains numerous papers on the development of new methods to treat effluents contaminated with dyes (Gupta et al. 2015; Punzi et al. 2015). However, biological treatments continue to be the most promising strategy: they are economically viable, efficiently reduce the organic load, and allow for treatment of large effluent volumes. Nevertheless, most microorganisms fail to degrade synthetic dyes completely, and they may generate aromatic amines that are even more toxic than the starting compounds (Aquino et al. 2014). Although physicochemical treatments are a good option from an economic viewpoint, they usually produce secondary compounds that can also be more toxic than the starting compounds.

In this context, advanced oxidation processes (AOPs) have emerged as a technology that can mitigate environmental problems via mineralization of toxic and harmful components, to give CO₂ and inorganic salts (Brillas 2014; Oturan & Aaron 2014). AOPs offer advantages such as high applicability and low energy costs, which make them sustainable.

The present study aimed to verify the efficiency of the simplest AOP and EAOP, namely the Fenton reaction and electrooxidation, respectively, in the treatment of an effluent containing the reactive dye basic blue 99 (BB).

Operational ease and the use of electron as reagent make the electrochemical technology a promising strategy. A metallic electrode coating can be used as catalyst, to form reactive species on the electrode surface and provide a more effective means to solve environmental issues related to industrial processes, particularly aqueous effluents (Särkkä et al. 2015).

During the electrochemical process, pollutants can be removed by either direct oxidation (when the pollutant is oxidized after adsorption over the electrode surface) or indirect oxidation (when the pollutant is oxidized by oxidants electrogenerated in the medium; e.g., HO^•, ⁠, and active chlorine) (Martínez-Huitle & Ferro 2006; Aquino et al. 2014).

This study employed dimensionally stable anodes (DSA^®) with commercial composition resembling the composition of the electrode used in the chloroalkali industry (Ti/Ru_0.3Ti_0.7O₂). These electrodes presented the following advantages: (a) interesting electrocatalytic properties and reduced anodic overpotential; (b) chemical and dimensional stability, which translated into better durability; (c) larger electrochemically active area; and (d) lower maintenance cost and energy consumption, among others.

The efficiency of the Fenton reaction and the electrochemical processes for the treatment of textile effluents is widely reported in the literature (Papadopoulos et al. 2007; Silva et al. 2011; Blanco et al. 2012; Solano et al. 2013; Nidheesh et al. 2013; Aquino et al. 2014). Most studies indicate the efficiency of both techniques in the removal of effluent color, but the total mineralization of the pollutant (dyes) is hardly achieved.

Papadopoulos et al. (2007) examined the effectiveness of Fenton oxidation for the reduction of the organic content of the wastewater generated from a textile industry in Athens (Greece). The experimental results showed that the maximum colour removal was 71.5% and the COD decreased by about 48% after 6 h of reaction.

Blanco et al. (2012) investigated the efficiency of the Fenton oxidation and the combined aerobic process and Fenton oxidation to degrade and reuse a real textile wastewater. The stand-alone Fenton process achieved 64% total organic carbon (TOC) reduction and >99% Escherichia coli removal. However, the best results were obtained when applying Fenton oxidation associated with the aerobic process. In this case 92% TOC and >99% E. coli removal were accomplished.

In electrochemical treatment of dye effluents, the efficiency is directly related to the composition of the electrode material, pH and temperature, but the highest influence is related to the presence or absence of chloride which considerably increases the efficiency of the process. Even in the presence of chloride, the total mineralization cannot be achieved. Studies carried out by Silva et al. (2011) on the degradation of the reactive dyes reactive blue 4 (RB-4) and reactive orange 16 (RO-16) showed that RuO₂ (DSA^®) electrodes were able to promote total color removal, at low chloride concentration (0.01 mol L⁻¹); however, only 80% COD removal was reached.

Solano et al. (2013) studied decontamination of real textile industrial effluent on diamond electrode (BDD). Results obtained in this research clearly demonstrated that the effect of the electrogenerated strong oxidant species, peroxodisulfates or active chlorine, depends on the electrocatalytic mechanism followed on the BDD surface, improving the color and COD removal. However, there is a limit to the NaCl concentration used for treating real effluents, to avoid the formation of organochloride compounds.

Despite the large number of studies on the degradation of dyes from aqueous solution by Fenton processes (Nidheesh et al. 2013) and electrochemistry, little attention has been paid to assess the ecotoxicity of treated dye effluent, when the chemical parameters of it meet the standard.

Ecotoxicity assays are an important means to assess the toxicity of the treated effluent. During effluent treatment, toxic intermediates may arise, or innocuous molecules may undergo bioactivation, so that the effluent displays higher toxicity than the untreated effluent (Pires & Chaparro 2010).

Lettuce seeds are excellent tools for effluent toxicity bioassays. They are highly sensitive to chemical stress and germinate fast, ensuring a rapid and reproducible assay. On the basis of bioassays that measured Lactuca sativa (lettuce) root growth and seed germination rate, this work also investigated how efficiently the Fenton reaction and electrooxidation diminished dye concentration, accelerated the remediation reaction, and decreased the ecotoxicity of the solution.

Therefore, the present study evaluates the efficiency of the Fenton reaction and of the electrooxidation process in the treatment of an effluent containing a contaminant dye and evaluates the phytotoxicity of the treated effluent.

All the reagents were commercially available. The dye BB was purchased from Pluryqumica, Brazil. Hydrogen peroxide (H₂O₂ – 50% m/m), ferrous sulfate (FeSO₄.7H₂O), sulfuric acid, and sodium chloride were supplied by Vetec. Sodium sulfate (Na₂SO₄) was acquired from ChemCruz. All the reagents were analytical grade and were used without purification.

The effluent containing the model dye BB (at a concentration of 0.075 g L⁻¹ or 0.75 g L⁻¹) was treated in a batch reactor containing 100 mL of the reaction solution. A continuous magnetic stirring system was employed, and the solution pH was maintained at 3.0 ± 0.1. The Fenton reagent concentrations were as follows: from 0.1 to 0.5 mmol L⁻¹ Fe²⁺ and from 10 to 50 mmol L⁻¹ H₂O₂, based on a 2³ experimental design. The H₂O₂, Fe²⁺, and BB concentrations constituted the experimental design input variables, i.e., factors that were evaluated at two levels, lower (−1) and upper (+1) levels (Table 1). All the experiments were conducted in duplicate. The reaction time was 1 h. Reaction efficiency was assessed by spectrophotometric analysis and toxicological tests.

All the experiments were performed in a 100-mL undivided open cell equipped with magnetic stirring. Commercially available DSA^® plates (Ti/Ti_0.66Ru_0.34O₂) with a surface area of 16 cm² were used as cathode and anode.

To determine the influence of experimental conditions for dye degradation in the model effluent, a 2³ factorial experimental design was employed. The influence of BB concentration (0.075 g L⁻¹ and 0.75 g L⁻¹), supporting electrolyte (0.2 mol L⁻¹ Na₂SO₄ and 0.2 mol L⁻¹ NaCl), and current density (10 and 30 mA cm⁻²) were evaluated in two levels, lower (−1) and upper (+1) levels (Table 1).

Solution samples were collected at regular times during the Fenton reaction and electrochemical process, and process efficiency was determined by spectrophotometric analysis. The maximum BB absorbance at λ = 578 nm was recorded on the spectrophotometer Shimadzu MultiSpec 1501 Series.

Ecotoxicological tests were accomplished by monitoring L. germination. L. sativa is recommended for this type of analysis because it is an important agricultural crop that is very sensitive to contamination by irrigation with residual water (Giorgetti et al. 2011).

L. sativa seeds were commercially available and presented high germination index (GI). Ten seeds were placed on filter paper impregnated with 5 mL of the treated effluent sample in Petri dishes. The untreated effluent was also used as reference. The negative control was distilled water. The assays were performed in duplicate. After 120 h, the number of germinated seeds, the stem length, and the root length in each plate were determined.

A low concentration of dye needs a small amount of hydroxyl radical, and parallel reactions with HO^• do not interfere in the effluent treatment. At high dye concentration the parallel reactions compete with the oxidation of the pollutant, affecting process efficiency.

Similar behavior was observed by Abo-Farha (2010) in studies of dye removal as a function of the concentration of Fenton's reagent and the concentration of the Acid Orange (AO) and Acid Red (AR) dyes. The maximum reduction of wastewater color was 85% and 80%, respectively, for AO and AR dyes using 10 mmol L⁻¹ H₂O₂ and 0.05 mmol L⁻¹ Fe²⁺. Further decrease in the dye decolorization in high concentrations of dye was associated to the formation of dimer molecules through a sequence of reactions from single dye molecules. Decolorization of the dimer molecule is more difficult, leading to the leveling off of color removal.

The chromophore group of the dye was not completely oxidized at better experimental conditions (0.5 mmol L⁻¹ Fe²⁺ and 10 mmol L⁻¹ H₂O₂), and 6% and 37% of the dye remained in solution in the presence of initial BB concentrations of 0.075 mg L⁻¹ and 0.75 mg L⁻¹, respectively. This probably resulted from formation of intermediate compounds that reacted with Fe²⁺, abstracted free iron from the solution, and formed stable iron complexes (Sires et al. 2007). The reaction of intermediate compounds with Fe²⁺, which inhibits the Fenton reaction, is common and often prevents the use of the Fenton reaction to treat some effluents. Alternatively, an associated oxidative process like electro-Fenton, photo-Fenton, or sono-Fenton is necessary to improve process efficiency (Babuponnusami & Muthukumar 2014).

In fact, in studies of Abo-Farha (2010) the decolorization efficiency was considerably higher for the photo-Fenton process, reaching the 100% reduction of solution color in the oxidation of AO and AR dyes.

A 2³ factorial design helped to evaluate how much each of the studied factors (BB, Fe²⁺, and H₂O₂ concentration) affected reaction efficiency. Table 2 presents the design matrix and the respective response (% of absorbance reduction) obtained for each experiment after 1 h of reaction.

Analysis of variance (ANOVA, Table 3) for the obtained statistical model revealed that the percent variation based on the regression, given by the ratio between the sum of squares due to regression and the total sum of squares (SS_R/SS_T), was 99.84%. This value should be compared to the maximum explainable figure, given by the difference between the total sum of squares and pure error divided by the total sum of squares ((SS_T−SS_pe)/SQ_T), which was 99.92%. Comparison of these values demonstrated that the linear model fitted well with the experimental data.

Another way to verify the good fit of the model was to compare the ratio between the means of the squares of the regression and residues (MS_R/MS_r) and the ratio between the mean of the lack of fit and pure errors (MS_lof/MS_pe) with the respective critical F. To state that the model fitted the experimental data, it was necessary to have test 1 (MS_R/MS_r>10×F) and test 2 (MS_lof/MS_pe<F).

For this model, the value obtained for test 1 was 1,324.15 > 35.80 (for 95% confidence), which evidenced a linear relationship between the factors and response of interest. The value obtained for test 2 was 3.45 < 5.32 (for 95% confidence), which confirmed fitting of the proposed model.

Evaluation of the relationship between the BB and Fe²⁺ concentrations showed that these factors significantly affected the response. Increasing BB concentration lowered the reaction yield. This effect was more pronounced when the experiment was conducted at lower Fe²⁺ concentration; that is, the process was more efficient at higher Fe²⁺ and lower BB concentrations (Figure 4).

The influence of the electrolyte on dye electrooxidation was also highlighted by Zhang et al. (2014). The studies show that, in the presence of chloride ion, the electrogenerated active chlorine played an important role in removing Acid Orange 7 (AO7) from synthetic dye wastewater using the Ti/RuO₂-Pt anode. No promotion of AO7 degradation efficiency was observed with the increasing of Na₂SO₄ dosage from 0.005 M to 0.05 M. When NaCl was present in the synthetic solution, the performance of Ti/RuO₂-Pt anode for AO7 removal was improved significantly and the removal efficiency increased with the increasing of the NaCl dosage. In the presence of 0.001, 0.005 and 0.01 M NaCl, the required time for complete degradation of 50 mg L⁻¹ AO7 was about 60, 30 and 25 min, respectively.

Silva et al. (2011) reported a similar result for the electrochemical oxidation of the dyes RB-4 and RO-16 in the presence of RuO₂ DSA. The electrodes degraded the dyes more efficiently in the presence of medium containing chloride, but organochlorinated compounds emerged as final products. The type of organochlorinated species resulting from the electrochemical reaction depended on the electrocatalyst composition. Ti/Ru_0.30Ti_0.70O₂ and Ti/Ru_0.30Sn_0.70O₂ were the most active electrodes for chloride evolution: they increased organic compounds oxidation and consequently the formation of organochlorinated species. The electrode Ti/(RuO₂)_0.70(Ta₂O₅)_0.30 was less active with respect to chloride evolution and therefore furnished lower amounts of chlorinated compounds.

Studies conducted by Solano et al. (2013) clearly demonstrated the effect of electrogenerated species like peroxodisulfates and active chlorine on the treatment of a real textile effluent. The authors verified that color and COD removal depended on the electrocatalytic mechanism taking place on the boron-doped diamond surface, and that a NaCl concentration limit had to be used during treatment of real effluents, to avoid formation of organochlorinated compounds. From an environmental standpoint, this result was extremely important because it increased the applicability of electrochemical processes in textile effluents remediation.

The use of NaCl also provided almost complete color removal within 5 min of reaction (Table 4). This prevented the study of the significance of effects only in the end of reaction because color removal as a function of time practically did not vary. This explained the decreased NaCl effect observed in Figure 6.

Considering the best experimental conditions for the Fenton reaction (0.5 mmol L⁻¹ Fe²⁺ and 10 mmol L⁻¹ H₂O₂) and the electrochemical treatment (30 mA cm⁻² in 0.2 mol L⁻¹ NaCl) (Figures 3 and 5), the electrochemical process was more efficient for BB oxidation than the Fenton process. In electrochemical treatment the BB absorbance reduced by 100% within a few minutes of reaction. For the Fenton reaction, maximum absorbance decreased by 94% and 63% at 0.075 mg L⁻¹ and 0.75 mg L⁻¹ BB, respectively.

Analysis of the kinetic parameters also attested to the higher efficiency of the electrochemical treatment.

Plotting of ln(Abs₀/Abs_t) as a function of k_ap t afforded a straight line, which indicated that the reactions followed first-order kinetics. The slope of the line provided the apparent global rate constant, k_ap, for BB oxidation. The electrochemical process and the Fenton reaction gave k_ap values of 1.1942 min⁻¹ and 0.05122 min⁻¹, respectively, in the better experimental conditions. The kinetic parameters could only be determined in the assays accomplished at higher dye concentration. The reaction at 0.075 g L⁻¹ was extremely fast, which prevented calculation of k_ap on the basis of the obtained data.

Concerning the kinetic aspects, the electrochemical process was 23 times faster than the Fenton reaction for BB oxidation in the studied conditions. Bearing in mind that ^•OH oxidized BB during the Fenton reaction, and that ^•OH has higher oxidation potential than free chlorine species, the higher efficiency of the electrochemical process could be associated with the amount of generated oxidizing species. Indeed, the quantity of oxidizing species produced during electrooxidation must have been significant as a result of the electrochemical properties of DSA toward the chlorine release reaction.

Reducing the level of toxicity of the treated effluent is as important as decreasing the physicochemical parameters. Many studies have used biospecies to evaluate effluent toxicity after remedial treatment (Valerio et al. 2007; Himanen et al. 2012).

Toxicity assays determine the toxic potential of a compound by measuring the response of a living organism. In the present case, lettuce (L. sativa) seeds were employed. Indeed, official organizations have recommended the use of L. sativa as the standard organism for toxicity tests.

Phytotoxicity tests based on L. sativa have many advantages: they are fast, simple, and reliable, and they do not require sophisticated equipment. In addition, L. sativa plants can be more sensitive to environmental stress than other test organisms (Valerio et al. 2007).

Toxicity tests were conducted only for the effluents treated in the better experimental conditions for the Fenton reaction and the electrochemical treatment. Table 5 displays the GI and the relative growth index (RGI) obtained for the seeds placed in contact with the treated effluents.

Before treatment, the effluent was considerably toxic, with 59.86% GI and 0.70 RGI. After treatment, both treated effluents presented higher toxicity. The effluent treated by the electrochemical process had 0% GI, whereas the effluent treated by the Fenton reaction presented 52.46% GI (Table 5, concentration 100%). The RGI also reduced to 0 and 0.426 for the electrochemical and Fenton processes, respectively.

The toxicity of treated effluent by the Fenton reaction is reported in the literature (Borba et al. 2013; Luna et al. 2014) and the toxicity depends on the pollutant, the experimental conditions and the organism used in toxicity tests. The Fenton reaction can reduce the toxicity of the effluent under optimized experimental conditions. For electrochemical processes, besides the possible formation of toxic product, the presence of the electrolyte contributes significantly to toxicity of the treated effluent. The presence of chloride, with the formation of hypochlorite, can or cannot increase the effluent toxicity. For microorganisms the active chlorine can be lethal, but for seed germination the chlorine can help in sterilization and break seed dormancy.

Evaluation of the acute toxicity as a function of the treated effluent dilution showed that L. sativa germination and growth inhibition decreased as the treated effluent became more diluted, regardless of the treatment process. Effluent dilution of 75% and 50% for electrochemical and Fenton treatment, respectively, led to lower toxicity as compared to the undiluted effluent before treatment. As for 12.5% dilution of the effluents treated by the Fenton reaction and by electrooxidation, these diluted effluents had significantly reduced toxicity: GI was 91.5% and 83%, respectively.

Garcia et al. (2009) conducted studies on the photocatalytic degradation of real textile effluents in TiO₂ and TiO₂/H₂O₂ systems. This author verified that the treated effluent affected L. sativa germination and growth less than the untreated effluent. The toxicity of the samples irradiated in the presence of TiO₂/H₂O₂ for 6 h was lower than the toxicity of the samples irradiated in the presence of TiO₂ only. This result indicated that more intense oxidation generated less toxic products.

Borba et al. (2013) also detected a small reduction in the toxicity of a tannery effluent treated by a photo-Fenton process. Despite the treatment, the effluent still presented elevated toxicity probably because recalcitrant substances remained at the end of the remediation process. A possible cause of the high toxicity level could be the nitrate concentration, which was not completely degraded by the photo-Fenton process. The authors suggested coupling the photo-Fenton process with other processes to treat industrial effluents.

Studies carried out by Luna et al. (2014) evaluated the ecotoxicity of five dyes to freshwater organisms before and during their photo-Fenton degradation for Daphnia similis, Pseudokirchneriella subcapitata and Ceriodaphnia dubia. Toxicity tests revealed that although the applied treatment was effective for decolorization of the dye, the partial mineralization may be responsible for the presence of degradation products, which can be more toxic than the original dye, as is the case of Vat Green 3 and Reactive Black 5, or lead to initially toxic products which may be further degraded to non-toxic products (AO7 and Food Red 17), or generate non-toxic products as in the case of Food Yellow 3.

Palácio et al. (2009) also verified elevated toxicity in a textile dyeing wastewater after 30 min of electrocoagulation (EC). The EC process alone was not able to remediate effluents satisfactorily, but it could be used as part of a complete effluent treatment system.

The Fenton and electrochemical processes effectively degraded the BB dye in a model effluent. Fe²⁺ concentration, current density, and supporting electrolyte were the most significant parameters during the process.

The Fenton reaction reduced BB concentration by a maximum of 94%. Residual BB removal was not possible because Fe²⁺ probably formed complexes with the reaction intermediates. The electrochemical process reduced the BB concentration by 97% within 60 min in the presence of Na₂SO₄ as supporting electrolyte. In the presence of chloride (NaCl as supporting electrolyte), BB concentration reduced by 100% within 5 min. These results demonstrated that electrogenerated chlorinated species influenced the oxidation process. However, studies to evaluate organochlorinated compound formation are desirable to validate process efficiency.

The kinetic and spectroscopic data indicate that the electrochemical process is more efficient for reducing the wastewater color, where 100% reduction of color was obtained within a few minutes of reaction compared with the Fenton reaction where the maximum reduction was 94% after 1 hour of reaction. Although the electrochemical treatment reduced effluent color more efficiently, the resulting treated effluent was more toxic to L. sativa germination and growth. The toxicity is possibly due to the presence of the hypochlorite formed. Studies are being conducted to decrease toxicity by eliminating the hypochlorite of treated effluent.

Hence, both the Fenton reaction and the electrochemical process can only be used to treat effluents containing BB when associated with other treatment processes.

The authors thank Coordination for the Improvement of Higher Education (CAPES), National Council for Scientific and Technological Development (CNPq – grant: 303630/2012-4), and Foundation for Research Support of the Alagoas State (FAPEAL) for financial support for this work.