Enhancement of E-Peroxone process with waste-tire carbon composite cathode for tinidazole degradation Open Access

Besides the enhanced generation of •OH, the E-Peroxone process may also reduce the formation of toxic byproducts. For example, Wang et al. found that the bromate formed in the E-Peroxone process was less than one-fourth of that in ozonation, probably due to the decomposition of O₃ and the reduction of hypobromous acid back to bromide by electro generated H₂O₂ (Wang et al. 2018). Thus, by simply installing electrodes in ozone contactors, the E-Peroxone process can significantly improve the removal of recalcitrant contaminants. Therefore, it is considered as a promising AOP thanks to the easiness in operation and scale-up and exclusion of chemical dosage as well (Zhan et al. 2016).

The electrode material plays a key role in E-Peroxone. Compared with the rare metal, carbon materials have been widely used as electrodes due to their characteristics such as the non-toxic, better chemical stability and corrosion resistance (Mounfield et al. 2018; An et al. 2019). The frequently used carbon materials as E-Peroxone electrodes are carbon nanotubes (Yang et al. 2018; Li et al. 2020) and activate carbon fibers (Zhan et al. 2016), and carbon/graphite felt (Liu et al. 2017; Yang et al. 2017a). The efficiency of the carbon materials can be further enhanced with polymeric substances like polytetrafluoroethylene (PTFE) and dimethyl silicon oil (DMS) (Xu et al. 2019). However, the above carbon materials are faced with the problems like high cost and non-green production due to the usage of non-renewable feedstock. Some studies have been conducted to make electrodes with waste-based carbon made from biomass (Huggins et al. 2014) and waste paper (Liu et al. 2012). However, to our best knowledge, waste-tire carbon (WTC), a carbon material made from the vast volume of waste tires, has not been studied as electrodes in E-Peroxone. The global increment of waste tires is approximately around 17 million tons per year now and would be up to 1.2 billion tons per year in 2030 s (Xu et al. 2021a). A potential solution for the waste tires is to produce carbon materials with the pyrolysis process. In addition, WTC also contains elements like N, O, S (San Miguel et al. 2003; Xu et al. 2021b), which are beneficial for the electrogeneration of H₂O₂ (Yang et al. 2017b; Ding et al. 2021).

Conversely, antibiotics are widely applied in clinics and animal husbandry to treat bacterial infections. Tinidazole (TNZ), a typical nitroimidazole antibiotic, is often used to prevent or treat amoeba infections, vaginal trichomoniasis and giardiasis (Qin et al. 2020). After administration, it could be excreted and enter the ecosystem via sewage, surface runoff, or soil infiltration. Consequently, TNZ has been detected in many water bodies including tap water, e.g. 1.8–17.8 ng/L in a report (Zhao et al. 2016). Researchers also found that TNZ may cause genotoxicity and cytotoxicity (Lopez Nigro & Carballo 2008). Therefore, it is vital to choose some efficient water-treatment technologies for the reduction of TNZ in the aquatic environment. TNZ is not highly reactive with O₃ (kinetic constant 330 M⁻¹ s⁻¹) (Rivera-Utrilla et al. 2010). Therefore E-Peroxone could be an effective alternative.

The objective of this study was to investigate, for the first time, the feasibility of modifying the graphite felt (GF) with WTC to enhance the performance of E-Peroxone in the degradation of TNZ as a target contaminant. PTFE was also applied to develop a composite electrode (WTC/PTFE-GF) whose generation efficiency of H₂O₂ was specially studied. Some operational parameters were investigated to identify the optimal condition. We also studied the degradation mechanism of TNZ, including the contribution of reactive oxygen species (ROS), the transformation products (TPs), and the degradation pathways.

WTC was provided by Rixin Hengli Rubber & Plastic Co. Ltd, and carbon black (Li-2060) was purchased from Cyber Electrochemical Materials Corporation. Tinidazole, PTFE (60 %wt.) and sodium indigotin-disulfonate were purchased from Aladdin; sodium chloride, sodium sulfate, anhydrous ethanol, potassium iodide, and tert-butanol (TBA) from Sinopharm Chemical Reagent Co., Ltd; sodium hydroxide, nitric acid, sodium thiosulfate, chloroform (CHCl₃), titanium potassium oxalate and sulfuric acid from Shanghai Lingfeng Chemical Reagent Co., Ltd. Except formic acid and acetonitrile being high-performance liquid chromatography (HPLC) grade, all other chemicals were at analytic grade. The deionized water was obtained from a pure water-treatment system (EPED-Z1-30T) and used during all the experimental processes.

The acidification of WTC was conducted by immersing 5 g WTC in 250 mL solution of 12 M HNO₃ for one hour. The round shapes of GF sheets with a diameter of 50 mm were firstly pretreated with a mixture of 30 mL acetone and 30 mL deionized water for 30 min. Then three pretreated GF sheets were separately immersed into the precursor mixture of 0.3 mL 60% PTFE emulsion, 60 mL ethanol and 1.2 g acidified WTC powder with carbon black powder (mass ration being 1:1), or 0.6 g acidified WTC powder with carbon black powder (mass ration being 1:1), or 0.6 g carbon black powder, named as 1.2 gWTC/PTFE-GF, 0.6 gWTC/PTFE-GF and 0.6 gCB/PTFE-GF, respectively. The three loaded GF sheets were heated in a muffle oven at 350 °C for 60 min. The loaded process was repeated several times until the loading was finally around 10 mg cm⁻² and no obvious bare surface of the cathode was observed. The anode was made of ruthenium–iridium titanium mesh supplied from Qinghe Yunxun Metallic Materials Co Ltd.

As shown in Figure 1, E-Peroxone reactions were carried out in identical cylinder reactors with an effective volume of 400 mL. Two electrodes (Anode: Ti mesh coated with Ru oxides; Cathode: WTC/PTFE-GFs) in the same shape (50 mm in diameter) were fixed in parallel at a distance of 15 mm in the reactor and connected onto a DC power source (ITECH, IT6721). Next, 50 mM Na₂SO₄ was used as an electrolyte except the test for chlorine effect in which Na₂SO₄ was partly replaced with NaCl to maintain the same conductivity.

The different gaseous ozone dosages were generated with a constant flow rate of 1 L min⁻¹ from a laboratory-scale ozone generator (Anseros, COM-AD) and was introduced via a diffusor at the reactor bottom. The degradation efficiency of TNZ was monitored along the reaction under various conditions, such as different ozone concentration, current density, initial pH values and electrolyte. The exhaust gas including O₂/O₃ was absorbed with a saturated solution of KI. To study the effects of ROS, varying dosages of TBA and CHCl₃ were applied as scavengers for •OH and ⁠, respectively.

The applied ozone concentrations were determined in aqueous phase by indigo disulfonate spectrophotometry at 610 nm. The concentration of H₂O₂ was measured using a potassium titanium oxalate method with a spectrophotometer (UV-1900PC, AOE Instruments, China) at 400 nm. The chemical oxygen demand (COD) was measured with a quick photometric method with potassium permanganate (China Standard GB 11892-89).

The TPs of TNZ were analyzed by UPLC-MS (Dinonex Ultimate 3000 UPLC) equipped with C18 column (100 mm × 2.1 mm, 1.9 μm) (Hypersil GOLD, Thermo Scientific, USA). The mobile phase was operated in a gradient mode at a flow of 0.2 mL min⁻¹ (20% B at 0 min., 100% B at 8 min., 20% B at 8.1 min.) where Channel A was water with 0.1% formic acid; and Channel B was acetonitrile with 0.1% formic acid. The sample injection volume was 5 μL. The UPLC system was coupled with a Thermo Scientific Q Exactive high-resolution mass spectrometer with electrospray ionization (ESI). Analytical parameters were determined in a positive ionization mode, acquiring spectra in a mass range between 50 and 750 m/z.

We analyzed the H₂O₂ concentration in 30 min with different cathodes, including 0.6 gCB/PTFE-GF, 0.6 gWTC/PTFE-GF and 1.2 gWTC/PTFE-GF, in 50 mM Na₂SO₄ solutions at pH 7.0, a value of most wastewater and water bodies. The current density was set as 10 mA cm⁻². Figure 2(a) illustrates the profiles of H₂O₂ concentrations. It can be seen that H₂O₂ concentrations in all tests were gradually increased up along the reaction time. The electrodes doped with WTC produced more H₂O₂ than that with CB at the same loading: 65 mg L⁻¹ vs. 23 mg L⁻¹ in 30 min. When the loading of WTC was doubled from 0.6 g to 1.2 g, the production of H₂O₂ was increased from 65 mg L⁻¹ and 86 mg L⁻¹ in 30 min. In addition, the corresponding current efficiencies of the 0.6 gCB/PTFE-GF, 0.6 gWTC/PTFE-GF and 1.2 gWTC/PTFE-GF cathode were 15.1%, 40.6% and 54.7%. Therefore, with the same dosage in the cathodes, the yield of H₂O₂ and the current efficiency with WTC were around 2.7 times of the data with CB. In Figure 2(b), we can find that the current efficiency was gradually dropping down along the reaction. This might be a result of the breakdown of excessive H₂O₂ at the cathode via a four-electron reaction (Yang et al. 2021).

The superior performance of WTC could be possibly attributed to the elements such as N and S in WTC which is absent in CB (San Miguel et al. 2003; Xu et al. 2021b). As shown in Figure 3, the modified WTC contained 0.39% S and 1.10% N. Nevertheless, further investigation are required to clarify the precise contribution of those elements.

Since H₂O₂ can be generated on the cathode in E-Peroxone and then ROS can be initiated, a better degradation of TNZ can be expected, as demonstrated in previous studies (Mao et al. 2018). As shown in Figure 4, when 20 mA cm⁻² current was applied, TNZ removal was dramatically enhanced at all studied ozone concentrations. For instance, at 11.3 mg L⁻¹ gaseous ozone, the removal efficiencies of TNZ in 15 min were increased from 99.5% to 99.90% and the corresponding first-order-kinetic constant increased from 0.35 min⁻¹ to 0.41 min⁻¹ when the current was switched on. As an essential reactant, ozone can effectively regulate the removal of TNZ in both processes. For example, in the E-Peroxone process, the increase of gaseous ozone from 7.8 mg L⁻¹ to 20.6 mg L⁻¹ elevated the removal of TNZ from 73% to 100% in 15 min, and the corresponding kinetic constant from 0.085 min⁻¹ to 0.78 min⁻¹. Nevertheless, it can be found that the enhancement of TNZ removal with electrochemistry was more significant at the low level of O₃ (e.g. 7.8 mg L⁻¹) than the high level of O₃ (e.g. 20.6 mg L⁻¹). This was probably because the removal of TNZ came from two reactions: O₃ and •OH formed from O₃/H₂O₂. When O₃ was increased, its contribution to the removal of TNZ can also be enhanced while the yield of H₂O₂ was likely to be constant at the same current density. Furthermore, the excessive O₃ may react with H₂O₂ and an ineffective decomposition of ozone may occur (Cornejo & Nava 2021). Therefore, it is important to match the yield of H₂O₂ and ozone dosage.

Current density is another important factor regulating the generation of H₂O₂ and subsequently may strongly influence the contaminant degradation in E-Peroxone. Figure 5 shows the effect of current density varying from 0 to 30 mA cm⁻² on the degradation of TNZ in E-Peroxone. When the high current density was applied, the degradation of TNZ increased correspondingly. This trend can be attributed to the fact that the high applied current density can enhance the electrogeneration of H₂O₂, which commonly happens in other studies (Valim et al. 2021). Consequently, more •OH could be produced to degrade TNZ at higher applied currents. As the generation of H₂O₂ is the key parameter in E-Peroxone, the low current density, 10 mA cm⁻² in this case, would generate a low level of H₂O₂ and thus result in an insignificant enhancement of TNZ removal in all dosages of ozone (Wang et al. 2018). By increasing the current density to more than 20 mA cm⁻², a catalytic effect can be clearly seen, especially at the gaseous ozone concentrations of 7.8 mg L⁻¹ and 11.3 mg L⁻¹. Nevertheless, when the gaseous ozone concentration was increased to 20.6 mg L⁻¹, the catalytic effect became negligible. The reasons should be similar to the discussion above on the results shown in Figure 4.

The energy consumption can be roughly estimated. Based on the scenario of 7.8 mg L⁻¹ and 1 L min⁻¹ ozone (468 mg h⁻¹), the energy consumption for ozone generation from air would be around 23 Wh (20 g kWh⁻¹ at the industrial scale). The generation of H₂O₂ (5 V, 30 mA cm⁻²) would consume 3 Wh, accounting for approximately 13% of the energy of ozone production. Nevertheless, when oxygen is used as a feed gas, the energy efficiency of ozonation generation could be further reduced.

The effects of initial pH values (pH 3-11) were investigated on the degradation of TNZ. To avoid the interference of inorganic ions, no buffer was involved in the tests. As shown in Figure 6(a), a higher removal of TNZ was obtained at higher pH values at the beginning stage (<3 min). Finally, a comparable removal efficiency was obtained after 5 min in all pH values. This may be attributed to the similar final pH values which were around 3.5 (Figure 6(b)). It is widely recognized that some acidic molecules can be formed during the oxidation of organic contaminants in the ozone-based AOPs leading the declination of pH to the range of 2.9–4.1 (Cornejo et al. 2021). As shown in Equations (5)–(9), the low pH value may impair the formation of ROS.

The trends of TNZ degradation with a function of its initial concentrations from 40 to 120 mg L⁻¹ were investigated with 50 mM Na₂SO₄, pH 7.0, the current density of 20 mA cm⁻², and gaseous O₃ dose of 10 mg L⁻¹. Figure 7 shows that the TNZ degradation occurred promptly and nearly complete removal of TNZ was achieved in 5–7 min. The corresponding apparent kinetic constants decreased slightly with the increase in initial TNZ concentration. The best kinetic constant of the TNZ degradation was obtained as the initial TNZ concentration of 40 mg L⁻¹: 1.37 min⁻¹ (R² = 0.96). The phenomenon might be related to the limited dissolved O₃. The high initial concentration of TNZ can consume more O₃ and the dissolved O₃ level would become limited when the mass transfer of ozone is constant. Moreover, the degradation of TNZ would generate TPs which could further complete for O₃ and ^•OH and then lead to a low degradation kinetic. For instance, after 5 min, there was still 89 mg/L COD in the test of 120 mg/L TNZ, several folds higher than the COD (24 mg/L) in the test of 40 mg/L TNZ.

In the oxidation processes, the organic contaminants are usually oxidized into some TPs before the complete mineralization. The TPs of TNZ in E-Peroxone were analyzed with the UPLC-MS system. According to the m/z data and the literature information (Acosta-Rangel et al. 2018), the molecular structures of five TPs can be estimated (Table 1), based on which of the degradation pathways were proposed. As shown in Figure 10, the degradation may possibly start with the hydroxylation on the N-heterocyclic ring with the formation of TP1. Further hydroxylation may continue to form TP2. The nitro group of TP1 can also likely be replaced with a hydroxyl group to generate TP3. Subsequently, the opening of the N-heterocyclic ring and the loss of methyl group could finally lead to TP4 and TP5.

In this study, we innovatively applied WTC as a recycled material to fabricate a cathode for the E-Peroxone process and compared its performance with commercial carbon black. The results demonstrated that the addition of WTC can significantly enhance the current efficiency of generating H₂O₂, up to 54.7%, which was about 2.7 times that of commercial carbon black at the same loading. For the efficient degradation of TNZ, the dosage of ozone should be fine tuned with the current density. In the study, the significant interference of chlorine (up to 50 mM) was not noticed. The main ROS were probably •OH and ⁠, as indicated by the scavenger tests. The degradation of TNZ may involve the hydroxylation and the opening of the N-heterocyclic ring, as the identified TPs indicated. In general, WTC can be a potential and economic material to replace commercial carbon for the construction of cathodes used in E-Peroxone. However, further investigations should be conducted to evaluate the long-term stability of the WTC cathode and to estimate the construction and operation cost.

The study was funded by the National Natural Science Foundation of China (Grant No. 52070097) and the Natural Science Foundation of Jiangsu Province, China (Grant No. BK20200699).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.