A H₂O₂-free heterogeneous Fenton process for the degradation of lincomycin using natural structural iron-containing clay mineral and dimethoxyhydroquinone with in situ generated hydroxyl radicals Open Access

Antibiotics are widely used for the treatment of human and animal infectious diseases and for promoting animal growth (Jin et al. 2019). However, these compounds have strong hydrophilicity and are not easily assimilated by organisms. Approximately 25–75% of unmetabolised compounds are discharged into the natural environment as faeces or urine (Chee-Sanford et al. 2001). Antibiotics have been detected in various environmental media, such as surface water, underground water, soil, sediment, and drinking water (Thiele-Bruhn 2003). Organisms are often exposed to antibiotic residues, such as veterinary antibiotics, due to the insufficient degradation of these contaminants (Bondarczuk et al. 2016). In addition, antibiotic residues may lead to ecological disturbance and resistant gene accumulation, posing a severe potential risk to the environment (Peng et al. 2017; Qian et al. 2018). Lincomycin (LCM) is a kind of lincosamide antibiotic that is commonly used in veterinary medicine and frequently detected in the environment. Different from most susceptible antibiotics, LCM is relatively persistent in the environment due to its chemical stability (Wang et al. 2012; Liu et al. 2016). The hydroxyl radical (^•OH) is an important reactive oxygen species with a standard reduction potential as high as 2.8 V (Wardman 1989); it can oxidise various organic contaminants at nearly diffusion-controlled rates (Yuan et al. 2018). The Fenton process is frequently applied in wastewater treatment. The key to the process is the reaction of Fe²⁺ species with H₂O₂ under strongly acid conditions to generate the highly reactive ^•OH. However, the traditional Fenton technology has two important shortcomings, including limited pH conditions, and secondary pollution of iron sludge (Ai & Hameed 2011; Zhu et al. 2019). To reduce the generation of ferric hydroxide sludge and circumvent the effect of limited pH range on homogeneous Fenton reactions, many researchers have focused on heterogeneous Fenton catalysis.

Clay minerals are important natural materials that have a wide variety of industrial applications because of their advantages, such as ubiquitousness, easy exploitation, low cost, and special physical and chemical properties. The role of clay minerals as catalysts in a heterogeneous Fenton reaction has also been extensively studied (Navalon et al. 2010). Compared with the traditional homogeneous Fenton reaction, clay-based Fenton reaction is also less limited by pH and produces notably less ferric oxide sludge. However, most previous studies have focused on Fe clay-based nanocomposites or Fe-exchanged pillared clay, which is prepared through the modification of clay minerals (Catrinescu et al. 2003; Feng et al. 2004). Many clay minerals contain Fe(III) in their crystal structure (i.e. structural Fe(III)) due to natural isomorphous replacement. For example, nontronite is a kind of clay mineral with high structural iron content (15–31%, wt%). However, their application to clay-based Fenton reactions has not received considerable attention. In addition, the method generally requires a constant supply of H₂O₂ (Herney-Ramirez et al. 2010), which will increase costs. Moreover, H₂O₂ is a hazardous substance that is toxic, explosive, unstable, and corrosive and will cause secondary pollution (Xiao et al. 2018).

However, in these studies, the Fe(II)-bearing clay minerals were artificially reduced to a great extent by iron-reducing bacteria or dithionite, a chemical agent with strong reducibility. In natural environments, these high-reduction-rate clay minerals are seldom exposed to an oxygen-rich environment. Meanwhile, the artificial reduction of these clay minerals is neither economical nor convenient for storage and transportation in engineering applications.

Hydroquinones are good electron donors and mediators due to their highly reactive electron transfer capability. These compounds can reduce Fe(III) minerals and generate Fenton reaction reagents (Varela & Tien 2003; Uchimiya & Stone 2006, 2009; Krumina et al. 2017). Hydroquinones are ubiquitous in both soil and aquatic natural organic matter. These compounds mainly come from the decomposition of organic litter material and microbial synthesis. In brown-rot wood degradation, 2,5-dimethoxyhydroquinone (2,5-DMHQ) was identified as the key fungal metabolite that takes part in the process (Varela & Tien 2003; Krumina et al. 2017). In the reaction, 2,5-DMHQ acts as a reductant towards ferric ions (Fe³⁺) to produce ferrous ions (Fe²⁺) and a superoxide radical (^•OOH). The superoxide subsequently dismutates and forms H₂O₂. Then, the reaction between Fe²⁺ and H₂O₂ can generate ^•OH to degrade wood (Rineau et al. 2012). These reactions give us a good indication for the construction of a clay-based heterogeneous Fenton process without addition of H₂O₂ and complex prereduction of clay minerals.

This study aimed to use natural clay minerals to degrade LCM through the production of ^•OH by 2,6-DMHQ, a commercially available isomer of 2,5-DMHQ. We hypothesised that this clay-based Fenton-like reaction is effective in degrading LCM without the addition of H₂O₂. Specifically, we aimed to address three questions: (1) Can ^•OH be produced in the nontronite/2,6-DMHQ system? (2) Can LCM be effectively degraded in this reaction system? If so, what are the reaction rate and the degradation pathway? (3) Can nontronite be recycled for sustainable production of ^•OH to degrade LCM?

LCM, 2,6-DMHQ, benzoic acid (BA), and p-hydroxybenzonic acid (p-HBA) were purchased from Sigma–Aldrich (MO, USA) with purities of >98%. High-performance liquid chromatography (HPLC)-grade acetonitrile, methanol, formic acid, phosphoric acid, and ammonia were purchased from Tedia Inc. (OH, USA). Hydrochloric acid (HCl), sodium hydroxide (NaOH), and isopropanol were purchased from Shanghai Chemical Reagent Inc. (China). The clay minerals of nontronite NAu-1 and NAu-2 were purchased from the Source Clays Repository of the Clay Mineral Society (www.clays.org). The clay minerals were homogenised Ca²⁺ of a specific size fraction (<2 μm). According to the unit cell formulas of smectites and our measurement, the Fe(III) contents (wt%) in the nontronite were approximately 24 and 26% for NAu-1 ((M⁺_1.0)[Al_0.58 Fe³⁺_3.38Mg_0.05][Si_7.00Al_1.00]O₂₀(OH)₄) and NAu-2 ((M⁺_0.97)[Al_0.52 Fe³⁺_3.32Mg_0.7][Si_7.57Al_0.01 Fe³⁺_0.42]O₂₀(OH)₄), respectively, with a small amount of Fe(II) [ (Wang et al. 2019).

As shown in Table 1, the molecule of LCM consists of a pyrrolidine ring, an amide moiety, and a pyranose ring (Calza et al. 2012). With the chemical formula of C₁₈H₃₄N₂O₆S, LCM has a molecular weight of 406.54 g/mol and, thus, a large molecular size (Liu et al. 2019). The pyrrolidine ring makes LCM an organic base, with a pKa of 7.97 at the amide N site. The pKa of the amide group and hydroxyl groups are −0.2 and >12.37, respectively. Hence, the dominating species of LCM is positively charged in this research, whose experimental pH fell below 7.97.

Table 1

Chemical properties of lincomycin (LCM)

.	Structure .	MW, g/mol .	Water solubility, mg/L .	logKow .	pKa .
LCM	(C18H34N2O6S)	406.54	92.19	0.29	Cationic < 7.97 Neutral > 7.97 and <12.37 Anionic < 12.37

.	Structure .	MW, g/mol .	Water solubility, mg/L .	logKow .	pKa .
LCM	(C18H34N2O6S)	406.54	92.19	0.29	Cationic < 7.97 Neutral > 7.97 and <12.37 Anionic < 12.37

View Large

The reaction mixture was around 200 mL suspension containing 1 g/L nontronite, 50 μM LCM, and 0.5 mM 2,6-DMHQ. The aqueous solutions were regulated at a desired value by 0.1 mM NaOH or HCl. The reaction was carried out at room temperature with a magnetic agitator. A total of 1-mL mixture was added at predetermined time intervals to 500 μL of 50 mM pH 11 isopropanol solution and extracted for 20 min. A preliminary experiment indicated that the recovery for LCM through this method was 97–101%. After extraction, the treated mixture was filtered by a 0.22 μm membrane. The residual LCM in the filtrate was analysed by HPLC (Waters Alliance 2965, Milford, MA). The degradation products of LCM were further determined by an HPLC system coupled with a high-resolution hybrid quadrupole time-of-flight mass spectrometer (HPLC-QTOF-MS, Triple TOF 5600, AB Sciex, Foster City, CA). Two control groups were set up for the clay mineral at pH 5 to ensure that LCM removal was truly due to the degradation by ^•OH: (1) LCM group to ensure that other factors do not cause a significant mass loss in the parent compound; (2) 2,6-DMHQ and LCM group to ensure the reaction between the two organics, which if present, have minimal effect on the overall results; and (3) LCM and nontronite group without the addition of polyphenol to ensure that LCM cannot be degraded by the clay mineral. Influencing factors, such as pH, initial concentration of LCM and recyclability of the catalyst, were investigated. Triplicate bottles were set up for all the experiments.

To measure ^•OH production, we set up parallel samples where LCM was replaced by ^•OH scavenger BA (10 mM). The product of the reaction between sodium benzoate, ^•OH, and p-HBA was measured to estimate the cumulative ^•OH concentration with a conversion factor of 5.87 (Liu et al. 2017).

The residual LCM was measured by HPLC. The HPLC system consisted of a Waters Atlantis T3 C18 column (5 μm, 4.6 × 250 mm) and an ultraviolet detector. The mobile phase was a mixture of 0.1% phosphoric acid (pH of 3.2–3.4 adjusted by ammonia) solution and acetonitrile (90:10, v/v) at a flow rate of 1 mL min⁻¹, with the detection wavelength at 210 nm. p-HBA was also quantified by HPLC. The mobile phase was a mixture of 0.1% formic acid aqueous solution and acetonitrile (65:35, v/v) at a flow rate of 1 mL min⁻¹, with the detection wavelength at 255 nm (Wang et al. 2021; Zhao et al. 2013).

A similar Atlantic T3 C18 column (Waters, 3 μm, 2.1 × 100 mm²) was used, and the mobile phase consisted of 0.1% formic acid (A) and methanol (B). The eluent gradient is listed in Table 2. The flow rate, column temperature, and injection volume were 0.2 mL min⁻¹, 30 °C, and 5 μL, respectively. Full-scan MS analysis was performed through positive-mode electrospray ionisation over the range of 100–1,000 (m/z). The source temperature was 550 °C. The atomising gas (GS1), heating gas (GS2), curtailing gas (CUR), temperature, declustering potential, and collision energy were set at 55, 55, 35, 550, 80, and 10, respectively. The ion spray voltage was set as 5,500 V. The MS data were analysed by PeakView software (Version 1.2, AB Sciex).

To further confirm the species of reactive free radicals, we examined the electron paramagnetic resonance (EPR) spectra of the reaction solution containing NAu-1; 2,6-DMHQ; and NAu-1 + 2,6-DMHQ. As shown in Figure 3(b), neither the single clay mineral nor the DMHQ system exhibited signal generation. By comparison, when NAu-1 and DMHQ coexisted, the obtained EPR showed four splitting peaks with an intensity of 1:2:2:1, and an EPR hyperfine fission structure of (N, NO) and (H, CH₂) = 14.9 G was observed, which were consistent with the spin parameters of 5,5-dimethyl-1-pyrroline N-oxide-OH radical (Peng et al. 2022). This result indicated that ^•OH was produced in the reaction process to induce the degradation of LCM.

HPLC-QTOF-MS was applied to identify organic products from NAu-1/DMHQ/LCM reactions. Table 3 summarises the proposed structures of the products based on the results and similar research. Similar to the degradation by potassium permanganate (Hu et al. 2011) and O₃ (Andreozzi et al. 2006), the sites of LCM where degradation starts include the aliphatic amine group on the pyrrolidine ring and the methyl thioether moiety. The pyrrolidine ring amine group oxidation results in the formation of iminium (I(a) and I(b)), alcohol (III(a) and III(b)), and amide (V and VI) products. The thioether group oxidation produces the corresponding sulphone (IV and VI) products. The sulphone product VI was also reported by Pospisil in their study of the oxidation of LCM by H₂O₂ (Posíšil et al. 2001). Another pathway observed in this reaction was the detachment of the hydroxyl group from the amide moiety, which produced a product with m/z 391 (II). This product (II) subsequently showed the detachment of the methylthio group and the formation of deprotonated m/z 344 (VII). This reaction pathway was also observed in the photocatalytic degradation of LCM on ZnIn₂S₄ (Gao et al. 2016).

Table 3

Products of LCM identified by HPLC-QTOF-MS

ID .	[M + H] +(m/z) .	RT (min) .	Molecular formula .	Difference from LCM .	Proposed structure .
LCM	407	3.5	C18H34N2O6S	/
I	405	2.8	C18H33N2O6S	–H
II	391	4.4	C18H33N2O5S	–OH
III	423	7.2	C18H34N2O7S	+ O
IV	439	3.1	C18H34N2O8S	+ 2O
V	421	9.5	C18H32N2O7S	–2H + O
VI	453	5.8	C18H32N2O9S	–2H + 3O
VII	344	8.6	C18H32N2O9S	–OH–SCH3

ID .	[M + H] +(m/z) .	RT (min) .	Molecular formula .	Difference from LCM .	Proposed structure .
LCM	407	3.5	C18H34N2O6S	/
I	405	2.8	C18H33N2O6S	–H
II	391	4.4	C18H33N2O5S	–OH
III	423	7.2	C18H34N2O7S	+ O
IV	439	3.1	C18H34N2O8S	+ 2O
V	421	9.5	C18H32N2O7S	–2H + O
VI	453	5.8	C18H32N2O9S	–2H + 3O
VII	344	8.6	C18H32N2O9S	–OH–SCH3

View Large

Based on the experiments results and similar research works, the reaction mechanism was proposed. Figure 4 shows the proposed pathways for LCM oxidation by NAu-1 in the presence of 2,6-DMHQ. The structural Fe(III) was firstly reduced by 2,6-DMHQ. Then, the formed Fe(II) polarised O₂ to produce ^•OH with the assistance of clay mineral surface. Meanwhile, LCM gradually adsorbed to the surface of the clay mineral and was attacked by ^•OH. The ^•OH reaction with the pyrrolidine ring amine group of the LCM molecule was initiated by the formation of iminium (products I(a) and I(b)). The hydrolysis of iminium and products III(a) and III(b) yielded hydroxylated products III(a) and III(b), respectively. III(b) was then oxidised by ^•OH to form the amide product V.^•OH reaction with the amide moiety producing the dihydroxyl product II. Demethylthio yielded product VII. The largest product signal was 1% of the parent compound, which indicated that most of the products were converted to CO₂.

^•OH can be produced from the oxygenation between structural Fe(II) and oxygen, providing a good idea for the development of green heterogeneous clay-based Fenton reaction. The most striking feature of this method is that it will not release considerable amounts of Fe ions, and the addition of H₂O₂ is unnecessary during the catalysing degradation reaction. However, previous studies have shown Fe(II)-bearing clay minerals with a high reduction ratio from artificial reduction by iron-reducing bacteria or dithionite, which is not economical nor convenient for the storage and transportation of clay minerals. Our study creatively applied 2,6-DMHQ to reduce the structure of Fe(III) in clay minerals and simultaneously degrade LCM, avoiding the problems of clay mineral transfer and anaerobic storage. In this application, iron-bearing clay minerals provided not only an iron source but also a constrained reaction environment in the galley regions for the highly efficient degradation of pollutants. Iron-bearing clay minerals can be effectively used in acid and neutral conditions and be efficiently reused without special treatment. Therefore, our method is environmentally friendly, efficient and simple and thus has a good application prospect.

This work was financially supported by the National Natural Science Foundation of China (grant 22006066) and the Open Fund from State Key Laboratory of Pollution Control and Resource Reuse of Nanjing University (PCRRF20038).

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

The authors declare there is no conflict.