https://orcid.org/0000-0001-6915-855X
Abstract
Antibiotic contamination in water has received significant attention in recent years for the reason that the residuals of antibiotics can promote the progression of antibiotic-resistant bacteria (ARB) and antibiotic-resistant genes (ARGs). It is difficult to treat antibiotics using conventional biological treatment methods. In order to investigate an efficient new method of treating antibiotics in water, in this study, microwave (MW) was employed in revitalizing peroxymonosulfate (PMS) to treat typical antibiotic tetracycline (TC). The Box–Behnken design (BBD) was applied to organize the experimental schemes. The response surface methodology (RSM) optimization was run to derive the best experimental conditions and validated using actual data. Moreover, the main mechanisms of PMS activation via MW were resolved. The results demonstrated that the relationship between TC removal rate and influencing factors was consistent with a quadratic model, where the P-value was less than 0.05, and the model was considered significant. The optimal condition resulting from the model optimization were power = 800 W, [PMS] = 0.4 mM, and pH = 6.0. Under such conditions, the actual removal of TC was 99.3%, very close to the predicted value of 99%. The quenching experiment confirmed that SO4•− and •OH were jointly responsible for TC removal.
HIGHLIGHTS
MW/PMS was employed to treat tetracycline.
BBD was applied to design the experimental protocol.
The equation between TC removal and influences was established.
RSM was used to predict optimal conditions.
TC was co-oxidized by SO4•− and •OH to CO2 and H2O.
INTRODUCTION
Global production of pharmaceuticals in treating human and animal diseases is increasing, which results in more pharmaceuticals being released into the aquatic environment during their fabrication and utilization (Aus der Beek et al. 2016). Generally, the wastewater from pharmaceutical plants is pretreated and then sent to a municipal wastewater treatment plant for combined treatment with municipal wastewater. However, the predominantly biological processes in municipal wastewater plants are not efficient in eliminating pharmaceutical components from pharmaceutical plants, and some of the pharmaceuticals are still discharged into the receiving waters (Khasawneh & Palaniandy 2021). In addition, most of the pharmaceuticals used by humans and animals cannot be completely metabolized, and some of their residues also enter water bodies through excretion. Antibiotics are common pharmaceuticals that are widely employed to treat diseases caused by various bacteria. Antibiotic residue in the aquatic environment can promote the expansion of antibiotic-resistant bacteria (ARB) and antibiotic-resistant genes (ARGs) (Jovanovic et al. 2021). Since the chemical composition and structure of most antibiotics are complex, some conventional water treatment methods (coagulation, filtration, adsorption, biotreatment) cannot effectively remove them (Wang & Zhuan 2020). Therefore, efficient treatment methods for antibiotics in water should be continuously developed.
) have been affirmed in treating recalcitrant organics like antibiotics in water (Duan et al. 2020). Typically, two oxidizing agents persulfate (PS, 
) and PMS (
) are applied in generating 
under specific catalytic conditions (Honarmandrad et al. 2022). Since the redox potential of 
(E0 = 2.5–3.1 V) is higher than that of PS (E0 = 1.96 V) and PMS (E0 = 1.75 V), sulfate radicals are more capable of degrading organic pollutants (Chen et al. 2018; Scaria & Nidheesh 2022). A number of substances or energies can be employed in activating PS or PMS to yield 
, such as transition metals (Gao et al. 2018; Bouzayani et al. 2019), carbon materials (Kohantorabi et al. 2021; Gao et al. 2022), heat (Norzaee et al. 2018; Ahn et al. 2021), ultraviolet light (UV) (Alayande & Hong 2022), ultrasound (US) (Gujar et al. 2023), and microwave (MW) (Qi et al. 2014, 2017). The above methods are effective in removing recalcitrant organic pollutants from water. The principles for producing sulfate radicals via these methods are summarized in reactions as shown in Equations (1)–(6) (Wang & Wang 2018). The disadvantage of the transition metal activation is that excessive metal ions will be introduced into the water, bringing secondary pollution to the water. The carbon material activation has the problem of difficult recycling of carbon materials. Compared to these two methods, energy-based activation does not require materials or chemicals and is therefore an environmentally friendly approach.
Heat activation usually refers to the heating of the reaction solution with conventional heating methods (water bath heating, electric furnace heating) to obtain sufficient thermal energy. As the applied thermal energy is higher than the bonding energy of the O–O bonds of PMS or PS, the O–O bonds are broken and thus free radicals are produced. UV activation of PS or PMS is also an energy-based approach. UV is a form of light energy that is converted to heat when it hits the PMS(PS)/contaminant solution to be treated, thus activating the PMS or PS. But unlike heat activation, UV also excites some of the water molecules, generating hydroxyl radicals (Song et al. 2022). US applied to the reaction solution can generate localized higher pressure (500–5,000 bar) and temperature (1,000–15,000 K) in the aqueous solution through cavitation, which will cause PS or PMS activation to produce free radicals (Gujar et al. 2023). Similar to UV, the US can directly cavitate water to generate hydroxyl radicals. MW activation (frequency from 0.3 to 300 GHz) is preferred among several energy-based methods due to its ability to rapidly warm aqueous solutions and efficiently degrade organic matter. In addition to its use in water and waste treatment, MW has been explored in other fields as well, such as waste management and treatment, phase separation and extraction processes, and remediation of soil (Bose & Kumar 2022; Shang et al. 2023). In this paper, we are more concerned with the application of MW in advanced oxidation methods for water treatment.
Some researchers have successfully degraded organic pollutants using MW-activated PS or PMS. For instance, MW/PS was employed to decompose sulfamethoxazole (Qi et al. 2014), methylene blue (KIm & Ahn 2014), pentachlorophenol (Qi et al. 2015), reactive Yellow 145 dye (Patil & Shukla 2015), humic acids (Zhang et al. 2018), imidacloprid (Genç & Durna 2018), sodium dodecyl benzene sulfonate (Bhandari & Gogate 2019), p-nitrophenol (Hu et al. 2019), dinitrodiazophenol (Wang et al. 2020), and tetracycline hydrochloride (Gao et al. 2020), and the influencing factors, kinetics, mechanism and acute toxicity were evaluated. MW has also been used as a coactivator for metal-activated PS, such as MW/zirconium oxide/PS degradation of methyl orange dye (Tantuvoy et al. 2023), and MW in assisting Cu-biochar to catalyze PS for the mineralization of oxytetracycline (Zhang et al. 2023). Furthermore, MW was applied to activate PMS in removing bisphenol A (Qi et al. 2017). MW was also used in combination with manganese ferrite to activate PMS for eliminating p-nitrophenol (Pang & Lei 2016). Feng & Li (2022) investigated the removal characteristics of 
and organic matter in the MW/PMS system with high Cl− content, and found that 
, 
, and 
played a major role in the removal of organic matter and 
in landfill leachate. Although MW/PS(PMS) processes have been reported in disposing of various recalcitrant organics, these processes have not been practically applied, thus more detailed studies are necessary. One aspect is that fewer studies of MW-activated PMS than activated PS have been reported, and no studies of tetracycline degradation by MW/PMS were found. Another aspect is that there is a lack of systematic optimization studies on the parameters of MW/PMS system in decomposing antibiotics.
In this study, MW/PMS process was applied to generate free radicals for typical antibiotic tetracycline (TC) decomposition. The entire data were analyzed using the software Design-Expert 13. The parameters of the experiment were set using Box–Behnken design (BBD) of the response surface methodology (RSM), the influencing factors were optimized depending on the resulting experimental data, and the dominant free radicals in MW/PMS system were also verified. Based on this study, the RSM method was used to optimize the reaction parameters of the MW/PMS system, which provided an effective method for predicting the removal rate of target pollutants.
MATERIALS AND METHODS
Chemical reagents
All chemical reagents used in this study were of analytical grade. TC and PMS (potassium monopersulfate triple salt) were purchased from Yuanye Biotechnology Co., Ltd (Shanghai, China). Tert-butyl alcohol (TBA) (C4H10O) was purchased from Maclean Biochemical Technology Co., Ltd (Shanghai, China), Methanol (CH4O) was obtained from Kelong Chemical Reagent Factory (Chengdu, China), sodium hydroxide (NaOH) was obtained from Bohao Chemical Co. Ltd (Leping, China), and sulfuric acid (H2SO4) was obtained from Beijing Chemical Reagent Factory (Beijing, China). The aqueous solution to be treated was prepared by adding chemical reagents to deionized water.
Experimental
A MW apparatus (IEC 60335-2-90, Analytik Jena AG, Germany) was used for heating TC solutions. The ablation tanks containing premixed TC and PMS solutions were placed in the MW apparatus and the reactions were initiated by powering up the MW apparatus. The pH was regulated with 0.1 M H2SO4 or NaOH as needed. The MW power and reaction temperature were controlled by function keys on the MW apparatus. Samples were extracted every 10 min and 0.5 mL methanol was added to each sample bottle and filtered via a 0.22 μm filter membrane. The treated water samples were put into a UV-visible spectrophotometer (T-U9, Youke Instrument Co., Ltd, Shanghai, China) to detect the TC concentration at 357 nm.
To assess the effect of 
and 
on removing TC, two radical scavengers, methanol (MeOH), and TBA were employed in MW/PMS/TC process.
Box–Behnken design
is the random error between predicted and observed values. The levels and independent variables are presented in Table 1.
The levels and independent variables
| Independent variables | Symbols | Range and levels | ||
|---|---|---|---|---|
| − 1 | 0 | + 1 | ||
| Power (W) | A | 600 | 700 | 800 |
| PMS (mM) | B | 0.2 | 0.3 | 0.4 |
| pH | C | 6 | 7 | 8 |
| Independent variables | Symbols | Range and levels | ||
|---|---|---|---|---|
| − 1 | 0 | + 1 | ||
| Power (W) | A | 600 | 700 | 800 |
| PMS (mM) | B | 0.2 | 0.3 | 0.4 |
| pH | C | 6 | 7 | 8 |
RESULTS AND DISCUSSION
BBD and analysis of variance
The coded and actual values based on BBD
| Run | A | B | C | TC removal (%) | |
|---|---|---|---|---|---|
| Actual value | Predicted value | ||||
| 1 | 700 | 0.3 | 7 | 82.3 | 81.9 |
| 2 | 800 | 0.3 | 6 | 89.3 | 89.9 |
| 3 | 700 | 0.3 | 7 | 81.5 | 81.9 |
| 4 | 700 | 0.4 | 8 | 91.1 | 91.9 |
| 5 | 800 | 0.2 | 7 | 78.3 | 78.5 |
| 6 | 600 | 0.4 | 7 | 90.8 | 90.6 |
| 7 | 700 | 0.3 | 7 | 81.4 | 81.9 |
| 8 | 700 | 0.2 | 6 | 87.8 | 87.0 |
| 9 | 600 | 0.2 | 7 | 90.5 | 88.8 |
| 10 | 700 | 0.3 | 7 | 80.4 | 81.9 |
| 11 | 700 | 0.2 | 8 | 73.5 | 75.7 |
| 12 | 700 | 0.3 | 7 | 83.9 | 81.9 |
| 13 | 700 | 0.4 | 6 | 93.4 | 92.2 |
| 14 | 800 | 0.4 | 7 | 95.5 | 97.2 |
| 15 | 800 | 0.3 | 8 | 86.2 | 83.7 |
| 16 | 600 | 0.3 | 6 | 88.4 | 90.9 |
| 17 | 600 | 0.3 | 8 | 87.0 | 86.4 |
| Run | A | B | C | TC removal (%) | |
|---|---|---|---|---|---|
| Actual value | Predicted value | ||||
| 1 | 700 | 0.3 | 7 | 82.3 | 81.9 |
| 2 | 800 | 0.3 | 6 | 89.3 | 89.9 |
| 3 | 700 | 0.3 | 7 | 81.5 | 81.9 |
| 4 | 700 | 0.4 | 8 | 91.1 | 91.9 |
| 5 | 800 | 0.2 | 7 | 78.3 | 78.5 |
| 6 | 600 | 0.4 | 7 | 90.8 | 90.6 |
| 7 | 700 | 0.3 | 7 | 81.4 | 81.9 |
| 8 | 700 | 0.2 | 6 | 87.8 | 87.0 |
| 9 | 600 | 0.2 | 7 | 90.5 | 88.8 |
| 10 | 700 | 0.3 | 7 | 80.4 | 81.9 |
| 11 | 700 | 0.2 | 8 | 73.5 | 75.7 |
| 12 | 700 | 0.3 | 7 | 83.9 | 81.9 |
| 13 | 700 | 0.4 | 6 | 93.4 | 92.2 |
| 14 | 800 | 0.4 | 7 | 95.5 | 97.2 |
| 15 | 800 | 0.3 | 8 | 86.2 | 83.7 |
| 16 | 600 | 0.3 | 6 | 88.4 | 90.9 |
| 17 | 600 | 0.3 | 8 | 87.0 | 86.4 |
ANOVA test for response function
| Source | Sum of squares | DF | Mean square | F-value | P-value |
|---|---|---|---|---|---|
| Model | 505.22 | 9 | 56.14 | 10.79 | 0.0024 |
| A – Power | 6.84 | 1 | 6.84 | 1.32 | 0.2890 |
| B – PMS | 207.06 | 1 | 207.06 | 39.82 | 0.0004 |
| C – pH | 55.65 | 1 | 55.65 | 10.70 | 0.0136 |
| AB | 71.40 | 1 | 71.40 | 13.73 | 0.0076 |
| AC | 0.7225 | 1 | 0.7225 | 0.1389 | 0.7204 |
| BC | 36.00 | 1 | 36.00 | 6.92 | 0.0339 |
| A2 | 69.92 | 1 | 69.92 | 13.44 | 0.0080 |
| B2 | 33.01 | 1 | 33.01 | 6.35 | 0.0398 |
| C2 | 12.89 | 1 | 12.89 | 2.48 | 0.1593 |
| Residual | 36.40 | 7 | 5.20 | ||
| Lack of fit | 29.58 | 3 | 9.86 | 5.78 | 0.0615 |
| Pure error | 6.82 | 4 | 1.70 | ||
| Cor total | 541.62 | 16 |
| Source | Sum of squares | DF | Mean square | F-value | P-value |
|---|---|---|---|---|---|
| Model | 505.22 | 9 | 56.14 | 10.79 | 0.0024 |
| A – Power | 6.84 | 1 | 6.84 | 1.32 | 0.2890 |
| B – PMS | 207.06 | 1 | 207.06 | 39.82 | 0.0004 |
| C – pH | 55.65 | 1 | 55.65 | 10.70 | 0.0136 |
| AB | 71.40 | 1 | 71.40 | 13.73 | 0.0076 |
| AC | 0.7225 | 1 | 0.7225 | 0.1389 | 0.7204 |
| BC | 36.00 | 1 | 36.00 | 6.92 | 0.0339 |
| A2 | 69.92 | 1 | 69.92 | 13.44 | 0.0080 |
| B2 | 33.01 | 1 | 33.01 | 6.35 | 0.0398 |
| C2 | 12.89 | 1 | 12.89 | 2.48 | 0.1593 |
| Residual | 36.40 | 7 | 5.20 | ||
| Lack of fit | 29.58 | 3 | 9.86 | 5.78 | 0.0615 |
| Pure error | 6.82 | 4 | 1.70 | ||
| Cor total | 541.62 | 16 |
Normal plot of the externally studentized residuals.
Actual values versus predicted values for TC removal.
, 
) during the period at pH 6–8 will affect the amount of free radicals produced, which in turn will affect the treatment of the pollutants. This phenomenon was similar to the results of a previous study on the degradation of bisphenol A using the MW/PMS process in a pH range of 2.70–8.18 (Qi et al. 2017). Figure 5 exhibited the pH and PMS concentration on TC removal. It is found that PMS concentration played a significant role in TC degradation. With PMS concentration increased from 0.2 to 0.4 mM, the TC removal rate obviously kept increasing. PMS, as an oxidizing agent, is a source of 
and 
, so an adequate supply of this agent promoted the formation of more free radicals, leading to more TC removed. 
The effect of PMS concentration and MW power on TC removal. (a) 3D; (b) 2D.
The effect of pH and MW power on TC removal. (a) 3D; (b) 2D.
The effect of pH and PMS concentration on TC removal. (a) 3D; (b) 2D.
RSM optimization and validation
Optimization of parameter values towards highest TC removal.
Free radical quenching experiment
(k = 1.0 × 107 M−1·s−1) and 
(k = 9.7 × 108 M−1·s−1), thus can quench both 
and 
. While TBA has a high rate constant for 
(k = (3.8–7.9) × 108 M−1·s−1) and a very low rate constant for 
(k = (4.0–9.1) × 105 M−1·s−1) and is therefore used to quench 
only. The impact of quenchers towards TC elimination is depicted in Figure 7. Compared with the control curve, the removal of TC at 60 min by the MW/PMS system decreased from 79.3 to 72.8% and 63.4% when 80 mM TBA and 800 mM TBA were injected, respectively. This phenomenon demonstrates that 
is formed during the reaction. When just the TBA was present in the MW/PMS/TC system, the TC removal dropped very quickly, from 79.3% at the control line to 34.2% at TBA of 800 mM and to 54.3% at TBA of 800 mM with 60 min of reaction. This confirmed that not only 
but also 
were generated in the MW/PMS/TC system. This result also coincided with the reaction mechanism of MW-activated PMS (Equation (6)). ![The effect of MeOH and TBA. Conditions: [TC] = 20 mg/L, pH = 6.4, [PMS] = 0.2 mM, Power = 600 W.](https://iwa.silverchair-cdn.com/iwa/content_public/journal/wst/88/11/10.2166_wst.2023.375/1/m_wst-d-23-00765f07.png?Expires=1736582136&Signature=BtIeyEomNz1xBrnRCEbThKxTfSVCa3Nb7G2t23vehrDFKmWMHfFoLGlJo4A5SRaZ6Pq0z5bQecU5iyCez9Nfhsu-sL0wOysBANzbI4W-06QINaQStHCuWXIB5-kWZ61mjbu9JDqrUR~5cECwWD3jj1Ji-zsW9ccXINbuI5QjMcnb36YTdZDICvpREUgwb-OZUFwDhtS~wgwr2~M6hnI6X0gsekDmtWR3EPYDMHoGvwJZh9J4vp19rS-AbqKAqGrDNg9M11y1ZWd-N-xfY8Jc7k7dBdlkigUzQT~Apg67lWgb9R4AnP4rV-AAYyGD3CF8fxhM8OgQW9Ps-qUCnFOwWQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
The effect of MeOH and TBA. Conditions: [TC] = 20 mg/L, pH = 6.4, [PMS] = 0.2 mM, Power = 600 W.
The effect of MeOH and TBA. Conditions: [TC] = 20 mg/L, pH = 6.4, [PMS] = 0.2 mM, Power = 600 W.
CONCLUSION
In this study, the performance and mechanisms of MW/PMS system were demonstrated for the decomposition of TC. The RSM was utilized to model and analyze the impact of independent variables on TC removal. The quadratic model obtained had a high R2 (0.9328), testifying that the model was reliable in predicting the experimental data. The optimum parameters predicted for TC removal (99%) were pH = 6, [PMS] = 0.4 mM, and MW powder of 800 W. The actual experimental result upon the optimal condition was very similar to the predicted result, proving that the developed model was very suitable for the MW/PMS system in this study. Also in terms of mechanism, it was concluded that 
and 
were produced in the MW/PMS process, which together acts on TC decomposition.
ACKNOWLEDGEMENTS
This work was supported by the Natural Science Foundation of Jilin Province of China, grant number YDZJ202201ZYTS681.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.
%20cropped.png?versionId=7151)