Solar activation of persulfate for simultaneous degradation of antibiotic drug and edible dye in a thin-layer flow photo-reactor

2.1. Materials

OTC (C₂₂H₂₄N₂O₉, 98.1%) was purchased from Huashu, and ERY (C₂₀H₆I₄Na₂O₅, 97.1%) from Alvan Sabet. Ferrous sulfate heptahydrate (FeSO₄.7H₂O, 99.5%), potassium PS (K₂S₂O₈, 99.0%), hydrochloric acid (HCl, 37.0%), sodium hydroxide (NaOH, 97.0%), tertbutyl alcohol (t-BuOH, C₄H₁₀O, 99.0%), ethanol (EtOH, C₂H₅OH, 99.0%), tetrahydrofuran (THF, C₄H₈O, 99.8%), 1,4-benzoquinone (BQ, C₆H₄O₂, 99.8%), calcium chloride (CaCl₂, 99.8%), sodium chloride (NaCl, 99.5%), calcium phosphate (Ca₃PO₄, 99.8%), sodium bicarbonate (NaHCO₃, 99.5%), and humic acid sodium salt (C₉H₈Na₂O₄, 99.1%) were Merck products. To adjust the pH of solutions, dilute solutions of hydrochloric acid and sodium hydroxide were used. Solutions were prepared with a qualified water (conductivity less than 0.08 μS/cm).

2.2. Analytical methods

The calibration curves of the OTC and ERY concentrations in the samples were first established using a UV-visible spectrophotometer (TEKSAN, LENA, DB-G100). To identify the intermediates, liquid chromatography-mass spectrometry (LC–MS) (Shimadzu LCMS 2010 A) analysis was utilized.

The corresponding UV–visible spectrum for the mixture of 20 mg/L of OTC and ERY (initial concentration), together with the chemical structure of the substrates have been presented in Figure 1. The distinct absorption peaks appear at 353 and 526 nm for OTC and ERY, respectively, consistent with previous reports [25,26]. Samples (2 mL) from the reactor content were withdrawn and the residual pollutant concentration was analyzed by measuring light absorbance relevant to calibration curves. The percentage of degradation efficiency (DE) was then obtained from:

Close

Figure 1.

The UV-visible spectrum and the chemical structures of the mixture of OTC and ERY in aqueous solutions; each 20 mg/L.

Export to PPT

in which $C_{\text{0}}$ and $C_t$ stand for the initial and a specific time concentrations for either of the pollutants.

2.3. Reactor set-up and procedure

The utilized setup has been schematically presented in Figure 2. The solar photo-reactor consisted of a brilliant stainless-steel plate, 300 cm² area, which was dot-stippled to make the whole surface available for the flow of aqueous solutions with no channeling. It was with a 20 ^° slope relative to the horizontal level and placed on a jack to adjust the distance from light source as well as the plate slope. A circulating pump receives solution from the bottom downcoming chamber and delivers to the top reservoir for feeding the plate just by passing its horizontal edge. Upon circulating, a thin-layer flow of about 3 mm was established from top of the plate to the below downcomer while the whole surface envisaged the irradiated solar light. The plate surface was gently polished with fine emery paper to remove deposits or discoloration from the brilliant stainless-steel plate and thus reflecting the passing light beams.

Close

Figure 2.

The schematic of the photo-reactor set-up consisted of the main reactor body and the plate of thin layer flow (1), simulated solar light irradiation source (2), jack to adjust the distance from light source and plate slope (3) and circulating pump (4).

Export to PPT

The reactor set-up was installed inside a standard solar simulator (Fan Daghigh Kowsar, SIM 2020) in which the light intensity of 1000 W/m², (measured by a solar power meter, CEM-DT 1307), was established. A corresponding standard for solar irradiance have been documented by the American Society for Testing of Materials (ASTM) based on the air mass (AM1.5, G173 global tilt data) where “1 Sun” corresponds to 1000 W/m² [27,28]. The light sources were three xenon lamps (300–1100 nm, 150 W), two monochrome LEDs (390 nm, 10 W), two LEDs (850 nm, 10 W), and one LED (940 nm, 10 W).

Preliminary experiments confirmed that by increasing light intensity from 50 W/m² to 300 W/m², relevant to different distances of the plate center from light source, the DE was increased. Indeed, the higher photon flux; the higher PS activation would be; however, light intensities more than 200 W/m² (at 15 cm distance from light source) would give no sensible change in efficiencies, apparently since most active radicals were previously produced. This matter has been explained in detail by Tromholt et al. [29].

In this study, the initial concentrations of OTC and ERY were 20 mg/L each, same as those previously utilized [30,3]. The lethal concentration (LC50) of OTC and ERY for microorganisms is less than 1 mg/L, which is than the allowed amounts entering surface water and soil. To run each experiment, 1 L of the solution was prepared, and a certain amount of PS was added once, followed by pH adjustment. Prior to each experiment, a 5-min warm-up was made for the lamps to stabilize the irradiation, and then the solution was transferred into the reactor. To follow the DE, samples (2 mL) were withdrawn at different times and were immediately analyzed to determine the concentration of each pollutant by measuring light absorption at the corresponding maximum wavelength by the UV-visible spectrophotometer. Experiments were conducted according to the predicted list of experiments and under a dominant room temperature of 22 ± 2 ^° C. Noteworthy, no sensible change in the temperature of solutions appeared during each experiment. Each experiment was repeated at least twice to ensure consistency.

2.4. Design of experiments

Aiming to reduce the number of experiments with respect to the considered parameters, the Central Composite Design (CCD) program, within the framework of response surface methodology (RSM), was employed for the design of experiments [31]. The star points in CCD were situated at the centers of each factorial space face, which gives fewer experiments at the levels of −1, 0, and +1 [32]. The PS dose, pH, and reaction time were the variables, and the average of the degradation efficiencies (${\text{DE}}_{\text{Av}}$) of OTC and ERY degradation efficiencies was considered as the system response, as previously introduced by Anjali and Shanthakumar [33]. The outlines, levels, and coded values assigned to the variables are listed in Table 1. A total of 28 experimental runs were proposed in the CCD model; they have been listed in Table 2.

The Design-Expert package (V.13) was employed to complete the design and the following data analysis. The relationship between independent variables (X) and the response (Y) was established in a quadratic response equation [33] (Eq. 7), in which ${\text{Х}}_{\text{i}}$ and ${\text{Х}}_{\text{j}}$ represent the coded variables, and ${\beta }_0$, ${\beta }_i$, ${\beta }_{ii}$, and ${\beta }_{ij}$, are respectively, the intercept, linear, quadratic, and interaction terms.

The analysis of variance (ANOVA) verifies the adequacy of the model for the prediction and the model significance related to the terms [34]. The coefficients of determination (${\text{R}}^{\text{2}}$ and ${\text{R}}_{\text{adj}}^{\text{2}}$) are determined. Particularly when a model is significant with a p-value of less than 0.05, the model equations are considered satisfactory to match the experimental data [35]. The qualified models have been presented in 3D surface plots.

3.1. The Influence of operating conditions

Results presented by Figure 3(a) show that DE rises with the PS concentration to about 180 mg/L and then tends to significantly decrease. As was indicated above, PS doses more than the optimal value could lead to conversion of sulfate radical anion species to sulfate anion, and thus lowering the oxidation of the target pollutants [36]:

Close

Figure 3.

Response surface and contour plots of the average DE of pollutants vs. (a) PS concentration and pH, and (b) degradation time and pH.

Export to PPT

The variation of the average DE versus the solution pH has been illustrated in Figure 3(a). The maximum level of efficiency has been revealed at a neutral pH of about 7. Very low and high pHs would decrease the DE. Indeed, under neutral pHs, the dominant species are mostly as ${\text{SO}}_{\text{4}}^{\text{•}-}$ species; however, under high acidic conditions, ${\text{SO}}_{\text{4}}^{\text{•}-}$ radicals are scavenged by themselves (self-quenching) and thus decrease the efficiency [37]:

At elevated pHs, on the other hand, the hydroxyl ion concentration increases with a potential of reacting with PS radical anion and hydroxyl radicals, producing less active species and reducing pollutants degradation [38]:

Evidently, Ouahiba et al. [39] reported the highest OTC degradation under the alkaline conditions of pH 9.1 in a UV/PS process. Bhandari et al. [40] reported the highest ERY degradation under acidic conditions (pH 5.1) in an ultrasound/PS process. Interestingly, here, a neutral pH was appropriate as a mediate value for degradation of the drug and dye pollutants.

Finally, Figure 3(b) depicts the influence of degradation time on the average DE. Higher efficiencies are achieved as the degradation time increases; however, no sensible change appeared after about 70 min. Thus, from the above findings, the optimum conditions in the Solar/PS process are introduced as: PS concentration of 180 mg/L, pH of neutral value 7, and degradation time of 70 min.

3.2. Modelling of the predicted results

To reach a comprehensive relationship for the average degradation of pollutants, the following quadratic model was appropriate:

The suitability of the model (${\text{R}}^{\text{2}}$= 0.9997 and ${\text{R}}_{\text{adj}}^{\text{2}}$= 0.9995, F-value of 5768.45, and p-value of less than 0.0001) denotes its significance [41]. Table 3 presents the ANOVA for the equation. The linear terms of pH (A), PS dosage (B), and degradation time (C) have individually a p-value of less than 0.0001, and with significant contribution in the introduced model for the average DE. According to the equation, a maximum ${\text{DE}}_{\text{Av}}$ value is predicted as 79.4% under the optimum conditions. A confirmatory experiment was accordingly performed, revealing 72.4% and 82.1% degradation efficiencies for OTC and ERY, respectively, i.e., an average efficiency of 77.3%, close to the predicted value by the model.

3.3. Adding ferrous ion

Figure 4 shows that adding ferrous ion as low as 1.5 mg/L effectively promotes PS activation so that 82.1% and 92.5% efficiencies were achieved for OTC and ERY, respectively, i.e., achieving an average DE of 87.3% after 70 min. The efficiency initially increases with the ferrous sulfate dosage and then decreases since excess amounts facilitate reaction between ${\text{Fe}}^{2+}$ cation and ${\text{SO}}_{\text{4}}^{\text{•}-}$ active species, and that ${\text{Fe}}^{2+}$ can rapidly convert to ${\text{Fe}}^{3+}$. Therefore, despite significant contribution in promoting PS activation, the efficiency could be limited via interaction of sulfate anion radical with excess amounts of ferrous ion (Eq. 5). Thus, an optimum 1.5 mg/L (0.026 mmol/L) was appropriate for the ${\text{Fe}}^{2+}$ dosage. The trend of the average degradation variation versus time for the Solar/PS/${\text{Fe}}^{2+}$system suggests a pseudo-first-order kinetics with a rate constant of 0.037 min⁻¹ corresponding to an R² value of 0.981.

Close

Figure 4.

The average DE versus degradation time with different ferrous ion concentrations under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

A comparative study on alternative processes, shown in Figure 5, reveals the important role of solar light and the ferrous ion in the activation of PS. As is obvious, their effectiveness in achieving higher average DE appeared in the sequence of Solar/PS/${\text{Fe}}^{2+}$ > Solar/PS >> PS/${\text{Fe}}^{2+}$ > Solar/${\text{Fe}}^{2+}$ > Solar.

Close

Figure 5.

The average DE versus degradation time for alternative processes, optimum conditions, and PS concentration of 180 mg/L and pH 7.

Export to PPT

A point to note here is that degradation of the OTC drug is stronger when accompanied by the ERY compared with the case of its degradation alone (Figure 6). It is since some sort of organic dyes (photosensitive materials) can be excited upon visible light irradiation and consequently assist PS activation to generate ${\text{SO}}_4^{•-}$ and ${\text{HO}}^{\text{•}}$ species. Solar light irradiation can excite electrons of these molecules from the highest level of occupied molecular orbital (HOMO) to the lowest level of unoccupied molecular orbital (LUMO), and the excited electrons can be joined to PS molecules to generate active species. Thus, dye molecules can act as “photosensitizer” when degrading in the process. Moreover, the photoexcited dyes are capable of oxidizing pollutants via electron abstraction, as is pointed out in a previous study (Saien and Jafari, 2022).

Close

Figure 6.

DE of OTC vs. time for the mixture and alone; Solar/PS/${\text{Fe}}^{2+}$ process under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

3.4. Influence of water coexisting matrices

There are a variety of water content matrices in water sources that can affect the DE of organic pollutants. Figure 7 represents the undesired effects of the presence of 0.026 mmol/L (the same ferrous ion concentration) of several water content ions of 1.6 mg/L bicarbonate (HCO₃^-), 2.4 mg/L phosphate (PO₄^3-), 0.6 mg/L sodium (Na ⁺ ‏), 1.0 mg/L calcium (Ca²⁺ ), and 5.2 mg/L humic (HA^-) in the Solar/PS process. For ease of comparison, no ferrous ion was included in this part of the experiments. The retarding effect of HCO₃^- could be relevant to its role in scavenging the sulfate radical anion, which leads to CO₃˙^- generation with limited activity [42]:

Close

Figure 7.

Variation of the average DE as a function of time in the presence of water coexisting matrices; all 0.026 mmol/L, Solar/PS process, and under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

Adding phosphate anion causes a decline in the average efficiency, from 77.3% to 46.18%, which can be related to the radical scavenging role of this anion, and that the ${\text{Fe}}^{2+}$‏ and ${\text{Fe}}^{3+}$ may form a precipitant compound with ${\text{PO}}_4^{3-}$, resulting in low availability for PS activation [42]. Further, non-significant effects (less than 10%) were revealed by adding ${\text{Na}}^{\text{+}}$‏ and ${\text{Ca}}^{\text{2+}}$‏ cations in their chloride salts, most probably related to the reaction between the accompanied chloride anion with active oxidizing species [43]:

Finally, the presence of ${\text{HA}}^{-}$, which acts as a natural organic substrate, competes with the pollutants in the degradation, causing DE to reduce from 77.3 to 62.0%, which is in agreement with a recent report [44].

3.5. Effectiveness of oxidizing radicals

Different quenchers of EtOH, t-BuOH, BQ, and THF, all with 3% v/v concentration, were used to deactivate sulfate, hydroxyl radicals, superoxide radicals (${\text{O}}_2^{\text{•}-}$) and singlet oxygen (${}_1\text{O}{}_2^{}$), respectively, under optimum conditions. As presented in Figure 8, the lowest ${\text{DE}}_{\text{Av}}$ values correspond to t-BuOH and EtOH, reaching 40.1 and 10.2%, respectively, compared to 87.3% for scavenger-free, which confirms the important role of hydroxyl and sulfate radicals. On the other hand, ${\text{DE}}_{\text{Av}}$ diminishes to low values by adding THF and BQ, indicating that singlet oxygen and superoxide radicals have a minor contribution in the degradation process. Based on these results, the reactive oxidizing species of hydroxyl radical, sulfate radical anion, singlet oxygen, and superoxide anion exhibit contributions of 47.2, 29.9, 12.8, and 10.1%, respectively.

Close

Figure 8.

The average DE as a function of degradation time under the influence of different scavengers (3%, v/v), Solar/PS/${\text{Fe}}^{2+}$ process, and under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

3.6. Comparison with previous studies

To represent the performance of the process in the utilized photo-reactor, a comparison was made with the previously reported homogeneous processes on the simultaneous photo degradation of pollutants. The operating conditions for the mixture of a drug pollutant either with a dye or with another drug, as well as the relevant conditions, have been listed in Table 4. As indicated, much less oxidant and ferrous ions have been consumed in the present study, while the achieved ${\text{DE}}_{\text{Av}}$ is significantly more (except for one study with UV irradiation), and the highest level of initial concentration of pollutants (40 mg/L) has been treated.

3.7. Intermediates and the degradation pathways

The mechanism, intermediates, and pathways of the pollutants degradation were investigated via LC–MS analysis. As presented by the spectra in Figure 9(a) and (b), there are two sharp peaks of the pollutants before treatment and ten intermediates after 70 min treatment (at m/z values of 737.1, 715.6, 559.2, 433.4, 428.2, 320.0, 302.1, 294.9, 128.0, and 105.0).

Close

Figure 9.

The LC–MS spectra of (a) pollutant solutions before and (b) after 70 min treatment with the Solar/PS/${\text{Fe}}^{2+}$ process under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

The most probable transition intermediates and pathways corresponding to these m/z peaks are depicted in Figure 10. With the action of ${\text{SO}}_4^{•-}$ and ${\text{HO}}^{\text{•}}$ prevailing oxidants, OTC degrades by alcohol group elimination, giving the intermediate identified with an m/z of 428.2, and then the amino group is oxidized to a hydroxyl group, giving another intermediate with the m/z of 320.9. Subsequent elimination of the hydroxyl group includes generating a product with m/z values of 302.9 and, in continuity, generating 2-ethylcyclohexanol and malonic acid with known m/z values of 128.1 and 104.0 [49]. Degradation of ERY, on the other hand, has been initiated by attacking oxidants on the polyaromatic rings, followed by debenzylation and decarboxylation reactions, giving an intermediate with the m/z of 737.1. This, in turn, can be decomposed into another structure (m/z of 715.6), known as 3-hydroxy-2,4,5,6-tetraiodoxanthen-1-one, and then it can break into another structure by eliminating iodide, giving a product with the m/z of 559.2. These intermediates can further degrade into other structures with m/z of 433.4 and 294.9. are known as, 4-benzyl-2,6-diodo cyclohexa-2,5-dien-1-one and 1-benzyl-4-iodobenzene.

Close

Figure 10.

The intermediates and proposed pathways for degradation of OTC and ERY based on the LC–MS analysis; Solar/PS/${\text{Fe}}^{2+}$ process under optimum conditions; PS concentration of 180 mg/L and pH 7.

Export to PPT

Based on the results, the employed process is effective for decomposing the structure of pollutants into simple compounds, most probably leading to ${\text{CO}}_{\text{2}}$ and light compounds at the end. Nearly similar intermediates and pathways have been proposed by Zhang et al. [49] and Liu et al. [50] in the photo degradation of OTC.

3.8. Toxicity evaluation

The standard, in vitro antibacterial activities of the initial (containing 20 mg/L of OTC and ERY) and treated solutions were determined by the antibiogram test (paper disc method) [51,52]. Two Gram-positive strains of Staphylococcus aureus and Bacillus subtilis, as well as two Gram-negative strains, including Escherichia coli and Salmonella paratyphi A were preserved at 4 ^° C and used. Then, 30 μg/disc of each solution was added to the petri dishes to determine the sensitivity for each of the four used bacteria. All plates were then covered and incubated for 18 to 24 h at 37 ^° C. The inhibition zones on each medium were determined by measuring the average diameter in millimeters observed from the lower surface of the petri dishes. All the experiments were carried out in triplicate. On each plate, an appropriate standard disc was applied depending on the antibiogram test bacterial strains from the Pad-Tan Company. Table 5 shows the results for the initial solution and the final solution treated by the Solar/PS/${\text{Fe}}^{2+}$ process under optimum conditions. As is apparent, the antibacterial activity was appropriate for the initial solution; however, the treated solution had no antibacterial effect, i.e., the treated solution had no effect on any bacterial strain.

According to the globally harmonized system of classification and labeling of chemicals (GHSCLC), the level of toxicities for the OTC, ERY, as well as the intermediates, were examined for fish, Daphnia, and green algae using the VEGA software (Version 1.3.18). The levels of acute and chronic toxicities of original pollutants and intermediates relevant to three aquatic organisms have been listed in Table 6. Compared with OTC and ERY, intermediates 2 and 6 (see Figure 10) are less toxic with both the acute and chronic toxicity criteria, whereas other intermediates exhibit a rather complicated behavior, i.e., their acute and chronic toxicities are scattered, and pollutants and intermediates have different toxicity levels with respect to the categories. Indeed, DE and the toxicity level of intermediates have to be considered together [53]. Generally, most of the intermediate toxicities have been reduced, and they exhibit less harm to the aquatic organisms than the original pollutants. Additionally, intermediates are found at very low doses with the effect of the process treatment, as indicated by the relative doses by the LC–MS analysis (Figure 9).