1 Introduction

Azo-type dyes are part of a group of pollutants that have been studied in various works. The objective of these studies is to promote the degradation of contaminants that are present in aqueous matrices and which originate from the effluents produced by the wastewater treatment plants of the textile industry (Peixoto et al. 2013). These compounds are characterized by the presence of one or more azo bonds (N=N) in association with aromatic and auxochrome systems (-OH, -SO3), which are the largest class of synthetic dyes in common use (Zanoni and Yamanaka 2016). When these substances are not properly removed from the effluent before being disposed of, they are a potential risk to human health and the environment, since they are considered toxic, mutagenic, carcinogenic, and non-biodegradable (Cai et al. 2016; Silva et al. 2017).

Humanity, therefore, confronts an environmental problem, and it is necessary to treat the contaminated water. Several conventional techniques, including physical, chemical, and biological processes, have been developed to remove azo dyes from the aquatic environment. These include adsorption (Rajabi et al. 2016), membrane filtration (Chen et al. 2015), coagulation/flocculation (Lau et al. 2014), and biological processes (Bonakdarpour et al. 2011). Most of these methods are not destructive and/or do not generate secondary pollution because the pollutants are transferred to another phase, which must be regenerated (Elhalil et al. 2016).

The treatment of textile effluents is so complex owing to the presence of compounds of low biodegradability and/or non-biodegradability. This is related to their intense color, variable pH, high salt content, and the organic load (Martins et al. 2018; Rodrigues and Külzer 2016).

Technologies based on chemical oxidation, such as advanced oxidative processes (AOP), have, therefore, been increasingly accepted as a viable treatment for wastewater containing biologically persistent and refractory compounds such as dyes (Baeissa 2016; Oller et al. 2011). AOP are capable of removing non-selectively dissolved organic pollutants, regardless of their chemical nature and the availability of dissolved oxygen (Pereira et al. 2016). They are considered clean technologies for the treatment of contaminated water, and their objective is to mineralize or decrease the recalcitrance of the more complex organic compounds and also to convert them into biodegradable or inert substances, such as carbon dioxide, water, and inorganic compounds (Araújo et al. 2016; Dewil et al. 2017).

Several studies have been carried out to evaluate the efficiency of AOP in the degradation of the dyes present in aqueous solution, such as those developed by researchers who have investigated the removal of the carmoisine dye from aqueous solutions using Fenton and photo-Fenton systems (Leite et al. 2016; Ertugay and Acar 2017; Nascimento et al. 2018). The aforementioned authors achieved a degradation of 92.7% when employing the Fenton process and of 95.1% with the photo-Fenton process. Lima et al. (2016) have evaluated the application of the UV/H2O2 process in order to degrade the 5G reactive blue textile azo dye. After 30 min of treatment, a discoloration in the order of 91.8% was observed, without a significant reduction in chemical oxygen demand (COD) (1.8%). A study on the degradation of Direct Blue 71 (DB71) using the Fenton process that removed 94% of the dye and attained an organic matter conversion efficiency (COD) of 50.7% after 20 min of reaction was carried out by (Ertugay and Acar 2017).

However, some researchers have been cautious as regards solutions/effluents submitted to AOP, since the formation of intermediates may be more toxic than the initial sample. The toxicity of aqueous solutions using dyes after treatment by AOP was evaluated by means of seeds of Lactuca sativa by Fernandes et al. (2018) and Elaissaoui et al. (2019). In their studies, Lima et al. (2016) performed ecotoxicity tests with Lactuta sativa seeds in aqueous solutions containing the reactive blue dye 5G before and after using the UV/H2O2 process, obtaining a considerable reduction in the phytotoxic levels. Santana et al. (2018) verified the phytotoxicity of solution contaminated by the direct orange dye 26 before and after submission to the Fenton and photo-Fenton processes. In the study, four different seed species were used: Lactuca sativa (lettuce), Cichorium intybus (chicory), Brassica juncea (mustard), and Portulaca grandiflora (eleven hours). From the results obtained, a higher sensitivity of lettuce seeds was observed. These authors, in addition to the seeds, also used bacteria in their ecotoxicological tests, after treatment of dyes via photo-Fenton process. It was observed, when using species of Escherichia coli and Proteus mirabilis, that the treatment used did not interfere in the development of the strains.

In this context, the general objective of this work is to propose a treatment using advanced oxidative processes to degrade the mixture of the direct black 22 and reactive red 195 textile dyes present in aqueous solution. The variables iron concentration ([Fe]), peroxide concentration ([H2O2]), and pH have, therefore, been studied in order to obtain the best working conditions in which to evaluate the kinetic study and toxicity. A mathematical model that is able to predict the degradation according to the variables [Fe], [H2O2], pH, and wavelength (λ) is also proposed.

2 Methodology

2.1 Chemicals

Commercial reactive red 195 (RR195) and direct black 22 (DB22) (Table 1) were used to prepare an aqueous solution to be studied through the degradation process by advanced oxidative processes (AOP). The solution of 25 mg L−1 of each dye was prepared with deionized water. In the degradation process, hydrogen peroxide (30% w/v, Química Moderna) and FeSO4∙7H2O (Vetec) were used as the iron source. To adjust the pH of the solution, it was utilized H2SO4 1 mol L−1 (Merck).

Table 1 Commercial RR195 and DB22 dyes
Full size table

Bench reactors with sunlight (Santana et al. 2017) and ultraviolet (UV-C: λ = 100–280 nm) (Zaidan et al. 2017) radiation were used in the reactions that employed radiation. The intensities of light during the experiments using sunlight and UV-C radiation were equal to 5.35 × 108 W cm−2 and 1.98 × 10−3 W cm−2, respectively, while the tests using solar radiation were carried out with a photon emission in the range of 3.9 to 7.4 × 108 W cm−2. It is important to note that both reactors are coated with aluminum foil.

The process of degradation was monitored through ultraviolet/visible (UV/Vis) spectrophotometer (ThermoScientific) with a linearity range of 2 to 100 mg L−1 for all the wavelength of maximum absorbance (λ): 544 nm RR195 and 476 nm DB22. It was determined the limits of detection and quantification of the method employed and also the coefficient of variance, according to INMETRO ( 2011). The same procedure was used to determine the aromatic groups (Paulino et al. 2015) observed for the dye studies, with values of λ equal to 218 and 290 nm.

2.2 Preliminary Study: Evaluation of AOP Efficiency

From the solution of the mixture of the RR195 and DB22 dyes containing 25 mg L−1 of each, it was studied the efficiency of the AOP: photolysis, Fenton, UV/H2O2, and photo-Fenton, during the time of 60 min. For each experiment, 50 mL of the aqueous solution was used. On the following experiments involving radiation, three types were employed: natural solar, artificial solar (sunlight), and UV-C.

In the Fenton and photo-Fenton processes, a concentration of iron [Fe] of 5 mg L−1 was used and the pH adjustment was performed for the range of 3–4. With the exception for the photolysis process, all the tests were performed using hydrogen peroxide concentration ([H2O2]) = 200 mg L−1. It is noteworthy that, throughout the research, H2O2 was fractioned in values equal to half of the total concentration of each test and added in the times of 0 and 15 min, according to a previous study by Tiburtius et al. (2009) who have found that the addition of this reagent fractionally increases the efficiency of the AOP.

2.3 Factorial Design

Based on the results of the preliminary study, the most efficient AOP was selected to be used during the study. In order to obtain a better working condition, a factorial design was settled up using three factors, in two levels, with analysis of the central point in triplicate, totaling 11 experiments. The studied variables were [H2O2], [Fe], and pH. Minimum and maximum levels of each variable were, respectively, [H2O2]: 100 and 200 mg L−1; [Fe]: 1 and 5 mg L−1; pH: 3-4 and 5–6. The duration of the experiment was 60 min, and the central level of the variables was determined based on the average of the minimum and maximum levels.

2.4 Evaluation of [H2O2]

With the results of the factorial planning, a univariate study of the [H2O2] was carried out, also with exposure to the radiation for a period of 60 min, maintaining the best conditions for [Fe] and pH. The [H2O2] studied were 40, 60, 80, 100, and 120 mg L−1, using 50 mL of the solution containing the dye mixture.

2.5 Kinetic Modeling

Based on the best experimental conditions of the AOP, tests were carried out to determine the kinetics of the degradation of the dyes studied, as well as the conversion of organic matter through the chemical oxygen demand (COD). In this study, 1 L of the solution containing the dye mixture was irradiated, aliquots were withdrawn at regular intervals of time for a period of 150 min. In this step, not only the wavelengths of maximum absorbance (λ) of the dyes (476 and 544 nm) but also the characteristics of the aromatic compounds (290 and 218 nm) were monitored. It was decided to evaluate these λ as they refer to the group of aromatic hydrocarbons which are responsible for conferring toxicity to the pollutants.

The kinetic modeling of the degradation process was performed according to a non-linear pseudo-first-order model. This model was proposed by Chan and Chu (2003), which is represented in Eq. 1.

$$ C={C}_0\cdot \left(1-\frac{t}{\rho +\sigma t}\right) $$
(1)

in which C is the concentration of the dye (mg L−1) after treatment at a reaction time t (min), and C0 is the initial concentration of the dye (mg L−1). The parameters ρ and σ represent the reaction kinetics (min) and the oxidative capacity of the system (dimensionless), respectively. The values of ρ and σ correspond to the linear and angular coefficients of the line, respectively, obtained from the linearization of Eq. 1 (Eq. 2).

$$ \frac{t}{1-\frac{C_1}{C_0}}=\rho +\sigma t $$
(2)

Then, the kinetic modeling evaluated the decay of the chemical oxygen demand (COD), using a first-order model (Rajkumar and Palanivelu 2004). This model considered the initial and final values of the parameter over time, according to Eq. 3.

$$ \ln \left(\frac{ CO D}{CO{D}_0}\right)= kt $$
(3)

where k is the velocity kinetic constant (min−1) for the conversion process.

It is noteworthy that COD was determined by spectrophotometric method. For this purpose, the procedure described in Method 5220D of the Standard methods for the examination of water and wastewater (Apha 2012) was followed.

2.6 Mathematical Modeling for the Prediction of Dye Degradation

A mathematical model was conducted using the Statistica 10.0 for forecasting of the percentage degradation according to the variables [H2O2], [Fe], pH and λ. In the first step, the effects table generated by the program for the variables involved was evaluated, getting a general degradation equation. Then, the profiles for predicting the values to understand the behavior of the variables were obtained.

In the second part, a simplification of the general equation was performed, observing which of the terms were statistically significant for a level of 95% confidence. Subsequently, in order to obtain a simpler equation, a new simplification was carried out, considering the percentage of degradation in different pH ranges.

2.7 Toxicity Tests

2.7.1 Toxicity Study with Seeds

The toxicity tests were performed considering two different organisms: seeds and bacteria. The seeds used were lettuce (Lactuca sativa) and clove of India (Syzygium aromaticum). For the seed, the method of evaluation consisted in exposing these to the solution of the dyes before treatment and after treatment for a period of 120 h at a temperature of 25 ± 1 °C in the absence of light. In these tests, 10 seeds were placed on filter paper in Petri dishes and moistened with 2 mL of different concentrations of the solution. The sample before degradation was not diluted, while the post-treatment solution was evaluated in the following concentrations: 100%, 70%, 50%, 10%, 5%, and 1%. All assays were performed in triplicate. Distilled water was used as a negative control and a 3% boric acid solution as a positive control.

At the end of the incubation, the seeds that showed germination were counted for each species, with subsequent evaluation of root growth. Based on this, the relative growth rate (RGR) and the germination index (GI) were calculated by means of Eqs. 4 and 5, respectively (Young et al. 2012).

$$ RGR=\frac{(RLS)}{(RLC)} $$
(4)
$$ GI= RGR\frac{(SGS)}{(SGC)}\times 100 $$
(5)

in which RLS is the total root length in the sample; RLC, the total root length in the negative control; SGS, the number of seeds germinated in the sample; and SGC, the number of seeds germinated in the negative control.

2.7.2 Toxicity Study with Bacteria

Pathogenic strains of Staphylococcus aureus (UFPEDA-02), provided by the Collection of Cultures of the Department of Antibiotics of the Universidade Federal de Pernambuco (UFPE), were selected to perform the bacterial toxicity experiment. Stock cultures of the bacterial strains were maintained at − 20 °C in sterile powdered skim milk containing 10% (v/v) glycerol. Bacteria were reactivated in Mueller Hinton Agar overnight at 36 °C and, subsequently, the colonies were resuspended in sterile saline solution (NaCl 0.15 mol L−1). The optical density at 600 nm (OD600) of that suspension was adjusted to the equivalent of 106 colony-forming units (CFU) per mL.

Subsequently, the percentage of bacterial growth was determined after incubation with the solution pre- and post-treatment and it was compared to the negative control (100% growth). This assay was performed in 96-well microtiter plates, where 40 μL of the Mueller Hinton broth medium, 80 μL of the sample, and 80 μL of the inoculum (106 CFU∙mL−1) were added to each well. Plates were incubated at 36 °C, and mean OD600 was measured between time zero and after 24-h incubation using a microplate reader. Each assay was performed in triplicate, and three independent experiments were evaluated.

3 Results and Discussion

3.1 Preliminary Study: Efficiency Evaluation of the AOP

In the first stage, in addition to photolysis, 3 different types of AOP were evaluated: UV/H2O2, Fenton, and photo-Fenton. It was found that the photolysis process regardless of the type of radiation employed (sunlight, solar, and UV-C) did not degrade the textile dyes under study. The UV/H2O2 process obtained very low percentages of degradation, presenting values lower than 4% when using sunlight and solar radiation at the characteristic wavelengths (544 and 476 nm). However, by employing UV-C radiation, it was possible to degrade between 30 and 46% of the dyes. This is due to the fact that, when using hydrogen peroxide together with some type of radiation, the indirect reaction occurs more energetically, considering that the oxidation potential of the hydroxyl radical (E0 = + 2.80 V) is higher than that of molecular hydrogen peroxide (E0 = + 1.78 V) (Brito and Silva 2012).

For the processes in which iron was used as a catalyst, it was found that the Fenton process was the least efficient, degrading around 65% for both dyes. The best results were obtained for the photo-Fenton/sunlight process, reaching 94.23% degradation for RR195 dye and 91.58% for DB22. It was also verified that the tests employing solar radiation had a behavior similar to that of the photo-Fenton/sunlight system, degrading 94.04% of RR195 and 84.09% of DB22, indicating that the solar radiation can be used with good efficiency and lower operating cost. However, it is known that the use of the photo-Fenton process under natural solar radiation has as a limitation the climatic changes associated with the seasons, that is, the presence/absence of emission of sufficient solar radiation to promote degradation of pollutants. Thus, more detailed studies were carried out using artificial sunlight.

These results are in agreement with the literature, since, when using radiation, a photo-reduction process of ferric ions occurs, allowing the regeneration of the ferrous ions, obtaining an increase in hydroxyl radical generation (Souza et al. 2010).

3.2 Factorial Design

The influence of the variables (factors) [H2O2], [Fe], and pH relating to the photo-Fenton/sunlight system was evaluated through factorial design. Statistical analysis and calculations of the main effects and interaction between the factors were performed using the Statistica 10.0 program. The factors were determined considering the significant effects at 95% confidence, as can be observed in the Pareto charts (Fig. 1a and b).

Fig. 1
figure 1

Pareto chart. (a) DB22/476 nm (pure error of 323.33 × 10−5) and (b) RR195/544 nm (pure error of 3.33 × 10−5); response surface for dye degradation: (c) and (d) [H2O2]-[Fe]; (e) and (f) [H2O2]-pH; (g) and (h) pH-[Fe]

Full size image

It was observed through the analysis of the Pareto charts (Fig. 1a and b) that both of the main effects and the interactions between them are relevant to the photo-Fenton/sunlight system. Thus, for a better understanding of the behavior of the factors, the response surfaces were generated for each of the dyes studied (Fig. 1c–h).

Based on the analysis of Fig. 1, it was possible to observe an analogous behavior for the degradation of the dyes under study. It was verified in both cases that the highest percentages of degradation were obtained using the lowest pH level (from Fig. 1e to 1 h))), which agrees with the negative values identified in the Pareto charts (Fig. 1). It was also possible to verify that, in the lower pH level (between 3 and 4), there is no influence of the concentrations of H2O2 (Fig. 1e, f))) and iron (Fig. 1g, h))), indicating that such reagents can be used in the smaller amount proposed in this study, although, combining the higher levels of these factors, a higher percentage of degradation was observed (Fig. 1e, f) )). The results obtained are in agreement with the literature, which state that the pH range between 3 and 4 is the best for the photo-Fenton/sunlight system (Dewil et al. 2017). According to these observations, [Fe] was fixed at 1 mg L−1 and the pH between 3 and 4; seeking an optimization for [H2O2] to be used, it has proceeded the univariate study of this variable.

3.3 [H2O2] Evaluation

It is known that the amount of H2O2 employed in the photo-Fenton process is of great importance, once this reagent must be present in excess to promote the degradation of the pollutants. Thus, the results of [H2O2] were evaluated as can be seen in Fig. 2.

Fig. 2
figure 2

Evaluation of the [H2O2] influence on the % of degradation

Full size image

Through the analysis of Fig. 2, it was found the variation between percentages of degradation (± 0.25% for RR195 and ± 0.16% for DB22) was not significant for [H2O2] values greater than 60 mg L−1. Therefore, two observations can be made: (1) the variation of the percentage of degradation is within the experimental error; (2) it is confirmed that, when the saturation of the ·OH radicals is reached, there is no favoring of the reaction, since the reagent is no longer consumed, being present in excess according to the study by Mitre et al. (2012). Thus, the degradation of the dyes studied showed an efficiency of over 99% using a [H2O2] equal to 60 mg L−1, [Fe] equal to 1 mg L−1, and pH between 3 and 4. For these conditions, it was performed the kinetic monitoring of the photo-Fenton/sunlight system.

3.4 Kinetic Modeling

With the best conditions found, the kinetic study was carried out by following the degradation of the chromophore groups responsible for the coloration, as well as the characteristic groups of the aromatic rings. For this, aliquots were withdrawn at times of 0, 25, 45, 65, 85, 100, 120, and 150 min.

Based on the experimental data of the time advance of the dye concentration for each λ analyzed, the kinetic model proposed by Chan and Chu (2003) was used. The kinetic curves obtained from such a model are shown in Fig. 3.

Fig. 3
figure 3

(a) Adjustment of the model to the experimental data of the temporal advance of the concentration of dye along the oxidation reaction; (b) comparison of the results of the residual fraction calculated versus experimental residual fraction (C/C0); (c) spectral monitoring of DB22 and RR195 dye degradation; (d) conversion data and kinetic parameters for dye modeling

Full size image

From the analysis of Fig. 3a), it can be stated that the proposed model fitted well to the experimental data, obtaining significant linear regression coefficients (greater than 0.9). The percentages of degradation obtained after 150 min, as well as the kinetic data derived from the application of the model proposed, are shown in Fig. 3d, being possible to observe that, in the first 40 min, there was a greater degradation rate, which is stabilized after 90 min.

According to Paulino et al. (2015), the higher the ratio 1/ρ, the higher the decrease in dye concentration. The ratio 1/σ is the maximum oxidation capacity at the end of the process. Then, the data in Fig. 3d) further show that, although the dyes under study, in their λ characteristic of color, exhibited an analogous degradation efficiency at the end of the treatment, DB22 showed the lowest decolorization rate (1/ρ), equal to 0.027 min−1 and a higher oxidative capacity (1.43). The same authors affirm that the different behaviors of the dyes can be associated to their molecular structure, being able to receive the influence of the quantity and arrangement of the aromatic rings. For a better visualization of the adequacy of the kinetic model, the graph was plotted relating the data of the proposed model to the experimental data (Fig. 3b)).

The comparison of the results of the calculated and experimental residual fraction (C/C0) shown in Fig. 3b) confirms that the proposed nonlinear pseudo-first-order kinetic model describes well the experimental results, which can be inferred once the closer the experimental values of the first bisector, the better the fit to the model. Finally, when analyzing Fig. 3c), the disappearance of the peaks in the visible region and the decay of the absorbance for the aromatic groups (λ equal to 218 and 290 nm) can be evidenced.

Thus, the COD behavior was evaluated over time and the organic matter conversion was analyzed. In Fig. 4, it is possible to observe the fit to the model referring to the experimental data of COD advance over time.

Fig. 4
figure 4

(a) Adjustment of the first-order kinetic model to the COD experimental data for an oxidation reaction of the dye mixture; (b) comparison of the results of the residual versus residual experimental fraction (COD/COD0)

Full size image

Based on the analysis of Fig. 4a), it was possible to verify that the proposed kinetic model fitted well to the experimental data, reaching a correlation coefficient in the order of R2 = 0.9927. The kinetic constant was obtained from linear interpolation, obtaining a value of k equal to min−1. Nevertheless, when analyzing Fig. 4b), it was verified that there was a good fit between the experimental residual fraction data and the model data, since the former are arranged around the bisector.

The use of the photo-Fenton/sunlight system in the treatment of the dyes solution promoted a reduction of the COD concentration on the order of 94.96% after 150 min. Although the chromophore group of the dye is destroyed by •OH radicals, some colorless degradation intermediates can be formed in the solution during the oxidation reaction, which justifies the non-proportionality of the reduction of COD against the high removal of coloring (Su et al. 2011). It is worth noting that the COD value for the solution before treatment was 46.26 mg L−1.

3.5 Equilibrium Modeling

In view of evaluation of DB22 and RR195 dye degradation by experimental design, a two-way ANOVA analysis was realized considering in the model both linear (L) and quadratic (Q) effects for [H2O2], [Fe], pH and λ. In Table 2 are shown the two-way ANOVA analysis results. The main effects Fe and pH were not significant for a quadratic model, once both factors are linear effects, whose combinations with other effects cannot be estimated for 95% confidence interval.

Table 2 Effect estimates by two-way ANOVA analysis
Full size table

According to model estimations, [H2O2] (Q) presented higher effect on degradation, following by [pH] (L), λ (Q), and λ (L). In this case, negative values were observed for [H2O2] (Q), pH (L), and λ (Q) contributing to reduction while λ (L) promoted degradation increase. Based on p values, [H2O2] by [Fe], [H2O2] by pH, and [H2O2] by λ are not significant (p value > 0.05) and can be removed from the model. Except in the cases of [H2O2] (L) and [Fe] (L), the main effects contributed more to the degradation model than two-way effects. In Fig. 5 are presented profiles for predicted values and desirability.

Fig. 5
figure 5

Profiles for predicted values and desirability: predicted profiles of H2O2 (a), [Fe] (b), pH (c), and λ (d); percental desirability (e) and desirability of H2O2 (f), [Fe] (g), pH (h), and λ (i)

Full size image

The profiles for predicted values and desirability indicated higher desirable degradation (Fig. 5e) for intermediate values of parameters H2O2, pH, and λ (red line). The effect of Fe on degradation (Fig. 5b and g) was the same for all concentrations investigated. Based on profiles of H2O2 (parabolic), Fe (constant), pH (decrescent), and λ (crescent) and disregarding experimental data for pH > 4.5 due to almost null degradation values, a new empiric model (model 1) was proposed according to Eq. 6.

$$ \mathrm{Degradation}\ \left(\%\right)=A\cdotp {\left[{\mathrm{H}}_2{\mathrm{O}}_2\right]}^2\cdotp \left[\mathrm{Fe}\right]\cdotp \ln \left(\lambda \right)/\mathrm{pH}+B $$
(6)

where A and B are constants. From the fit of experimental data to the model (least squares as loss function) was observed R2 = 0.32, resulting in an unsatisfactory prediction. In sequence, another empiric model (model 2) was proposed according to Eq. 7.

$$ \mathrm{Degradation}\ \left(\%\right)=A\cdotp {\left[{\mathrm{H}}_2{\mathrm{O}}_2\right]}^{k1}\cdotp {\left[\mathrm{Fe}\right]}^{k2}\cdotp {\mathrm{pH}}^{k3}\cdotp {\lambda}^{k4}+B $$
(7)

where A, B, and ki (i = 1, 2, 3, 4) are constants. In Table 3 is shown the fit of experimental data to the model 2 (least squares as loss function) at 95% confidence interval.

Table 3 Fit of experimental data to model 2
Full size table

The results indicate good fit of experimental data to model 2 presenting R2 = 0.98. The use of model was restricted to levels 100.0 mg L−1 ≤ [H2O2] ≤ 200.0 mg L−1; 1.0 mg L−1 ≤ [Fe] ≤ 5.0 mg L−1; pH ≤ 4.5; and 218 nm ≤ λ ≤ 290 nm.

3.6 Toxicity Test

3.6.1 Toxicity Study with Seeds

After subjecting the dye solution containing DB22 and RR195 to the photo-Fenton/sunlight process, the toxicity of the substances against the seeds of lettuce (Lactuca sativa) and clove of India (Syzygium aromaticum) was analyzed. Initially, it was performed the count of the seeds that germinated in the negative (water) and positive (boric acid) control, following for those ones that germinated in the dye solution before and after treatment (at different concentrations). The results of the germination averages are shown in Table 4.

Table 4 Average seeds germinated, root growth, RGR, and GI (%) of the Lactuca sativa and the Syzygium aromaticum seeds (threefold trials)
Full size table

From the analysis of Table 4, it was observed that the solution of dyes after the treatment behaved similarly to the negative control for both seeds. It is possible to observe that the number of germinations occurred inversely to the concentrations of the treated solution. In this context, it was possible to calculate root mean values and standard deviations of the radicles of each experiment, as well as the relative growth rate (RGR) and germination index (GI) in percentage.

According to the results presented in Table 4, it was found that, as the concentrations of the solution rose, the seeds did not undergo a significant inhibition in the growth of their radicles. It is also observed that, although the application of the AOP did not interfere in the germination, this process caused alterations in the development of the studied species, causing them a delay in the growth.

According to the literature, considered inhibition of growth are all results that are less than 80% (Young et al. 2012). Based on this and according to the results presented in Table 4, it was verified that the treated solution showed toxicity from a concentration of 10% for both seeds, since the RGR values were equal to 90% and 81% for the Lactuca sativa and Syzygium aromaticum seeds, respectively. When the industrial effluents are released into the water courses, it is diluted, which justifies the study of various concentrations of the treated solution. Thus, after submission to the photo-Fenton/sunlight process, there are no compromises to the life forms present in the environment.

Results from previous studies corroborate that the treatment of effluents through AOP causes a reduction in the toxicity of the samples. This fact was studied by Lima et al. (2016) in the degradation of dye after treatment with AOP, as well as by Palácio et al. (2012) after treatment of textile effluent through the photo-Fenton process, and by Napoleão (2015) in the treatment of pharmaceutical drugs.

3.6.2 Toxicity Study with Bacteria

Toxicity analysis was performed with the initial solution (SBT) containing the dyes studied and the post-treatment solution (SPT) in concentrations ranging from 1 to 100%. To evaluate the bacterial activity of the solutions, growth tests against S. aureus bacteria were carried out. Thus, the percentage of growth of the inoculated bacteria together with the C/2.5 dye solutions was considered and the analysis was performed by optical density at 600 nm (OD600). The results of the mean and the standard deviation (σ) of the replicates of the variation of the OD600 values between the initial time and after 24 h, as well as the percentage of growth of each treatment in relation to the negative control (100%), are set out in Table 5.

Table 5 Growth percentage of Staphylococcus aureus
Full size table

Based on the results shown in Table 5, it can be stated that the toxicity analyses using microbiological techniques proved to be viable for the environmental study and also to quantify the bacteria present in the sample (aqueous solution, effluent, river water). It was found that SPT presented better results than SBT, since the latter inhibited the growth of 71.1% of the bacteria, indicating a high toxicity. After the treatment, the percentage of inhibition was reduced considerably, passing to 46.0% for the sample without dilution and inhibiting only 33.4% for SPT1%. In order to evaluate the quality of water and effluent, tests with microorganisms are important, since the presence of certain microorganisms directly interferes with the quality of water for human consumption and use (Zaidan et al. 2017).

4 Conclusion

The photo-Fenton process under sunlight irradiation can satisfactorily degrade red reactive 195 and direct black 22 present in aqueous solution, removing the chromophoric (99%) clusters and reducing the concentration of persistent aromatic groups (> 76%). The study also revealed that the treatment follows nonlinear degradation kinetics, capable of reducing the organic matter and COD levels. Phytotoxicity studies using seeds of Lactuca sativa and Syzygium aromaticum showed that treatment with AOP did not inhibit the germination of these species; however, it led to a decrease in root growth, indicating a low toxicity of the post-treatment solutions. This fact was corroborated by the toxicity analysis using microbiological techniques, in which there was a reduction of the toxicity of the dyes after the submission to the process.