Insights into the degradation of doxycycline in water using electrochemically prepared TiO₂: Effect of environmental factors, reaction mechanism, degradation pathway, and toxicity evaluation

2.1. Synthesis and characterization of materials

DOX (C₂₂H₂₄N₂O₈) was obtained from Kunshan (China) and employed as the primary target compound for the photocatalytic degradation tests. Potassium chloride (KCl) from GHTech (China) was used as the electrolyte, and titanium metal bars (industrial grade) were used as the electrodes. Titanium (IV) oxide (P25, TiO₂, ≥ 99%) was purchased from Merck (Germany). NaOH, H₂SO₄, and HNO₃ from Xilong (China) were utilized to adjust the pH of the solution. Other chemicals included di-ammonium oxalate monohydrate ((NH₄)₂C₂O₄.H₂O, DiA) from Shanghai Zhanyun (China), sodium azide (NaN₃, NaN) from Himedia (India), and tert-butanol ((CH₃)₃COH, TeB) and p-benzoquinone (C₆H₄O₂, pBZ) from Merck (Germany).

The electrochemical synthesis of TiO₂ was conducted using titanium bars as anode and cathode, immersed in a potassium chloride electrolyte solution (Fig. S1). A 60 V/3 A DC power source (TES 6102) was employed, with the voltage gradually increased from 0 to 25 V while maintaining a current of 1.7 - 2.0 A. The reaction was stirred at a speed of 250 rpm and kept at a temperature between 40 - 45°C for 1 h. After that, the TiO₂ material was collected through vacuum filtration using a polyvinylidene difluoride (PVDF) membrane filter (0.2 µm pore size, 47 mm diameter). The filtered material was thoroughly washed with distilled water until the pH reached neutral (e.g., pH = 7). After washing, the material was placed in the oven to be dried at 80°C for 24 h and finally calcined at a temperature in the range of 300-700°C for 2 h to obtain the TiO₂ photocatalysts.

Scanning electron microscopy (SEM) was employed to investigate the surface morphology of the TiO₂ material, and its elemental composition was determined through energy-dispersive X-ray spectroscopy (EDX, JSM-IT200 instrument, JEOL, Japan). Fourier-transform infrared spectroscopy (FTIR, FT/IR-6X spectrometer, JASCO, Japan) was employed to identify the chemical bonds and functional groups on the surface of synthesized TiO₂. The crystalline structure of the TiO₂ was investigated using X-ray diffraction (XRD) using a D2 Phaser diffractometer (Bruker, Germany). The optical properties of the TiO₂ catalyst were examined through a U-3010 spectrophotometer (Hitachi, Japan).

2.2. Photocatalytic experiments

The DOX degradation experiments were conducted using a photocatalytic reactor system (Fig. S2). The main component includes a reactor for containing synthetic wastewater prepared with a predetermined concentration of DOX. To activate the photocatalyst for the degradation process, an 8W UV black lamp tube (with a wavelength of 365 nm) from Haichao (China), was used. The DOX concentration was measured using a UV-Vis spectrophotometer (model DR6000, Hach, USA).

In a typical test, TiO₂ photocatalyst (0.4 g) was weighed and subjected to ultrasonication for 30 min in water to ensure uniform dispersion. At the same time, 0.02 g of DOX was precisely weighed and introduced into the water, maintaining continuous stirring for 30 min. Meanwhile, the solutions of TiO₂ and DOX were mixed, followed by the gradual addition of DI water until the final volume in the reactor reached 2000 mL. Then, the pH level was adjusted to set a desired value (e.g., from 3 to 11). The mixture was then transferred into a sealed cylindrical reactor. Subsequently, it underwent stirring in the absence of a light source (dark-phase stirring, DPS) for 30 min. This process ensured adsorption equilibrium between the DOX and the TiO₂ catalyst.

Upon reaching adsorption equilibrium, the UV lamp was switched on to initiate the photocatalytic process. Continuous aeration was employed to maintain oxygen diffusion throughout the process. Samples were collected at specific time intervals of 5, 10, 15, 20, 25, 30, 45, 60, and 90 min and underwent centrifugation at 5000 rpm to remove the solid material. The liquid was then filtered through a 0.45 μm polyethersulfone syringe filter to remove residual particulates. Measurement of DOX concentration was performed using a UV-Vis spectrophotometer at an absorbance wavelength of 346 nm. The degradation efficiency of DOX was determined using the following equation:

Where C₀ denotes the initial concentration of DOX, and C_t represents the concentration at a specific time t, both are proportional to the optical absorbance measured by a UV-Vis spectrometer.

To verify the reduction of “DOX concentration-over-t” whether it follows the pseudo-first-order kinetic assumption, the plot of C_initial versus C_t and (or C₀/C_t) could be expressed by the natural logarithm as ln(C₀/C_t). If the reduction rate of DOX was fitted by exponential (for C₀/C_t) or linear growth (for ln(C₀/C_t)), the degradation process can be defined by using a pseudo-first-order kinetic model (Eq. 2) or Langmuir-Hinshelwood (L-H) model (Eq. 3) (Nasiri et al., 2019):

Where k_obs is the apparent reaction rate constant (min^-1), K_C is the surface reaction rate constant (mg/(L.min)), and K_L-H is the adsorption equilibrium constant (L/mg) (Malakootian et al., 2020). K_app (min^-1) represents the apparent rate constant, influenced by both adsorption and reaction kinetics. The model parameters, particularly K_app and K_L-H, were estimated using time-resolved DOX concentration data.

Additionally, the formation and the role of ROSs on photocatalysis were examined by introducing selective radical scavengers. Liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) was utilized for the identification and characterization of the degradation products during the photocatalytic reaction using an Acquity UPLC I-Class/Xevo TQ-S Micro system (Waters Corp., Milford, USA). In addition, ECOSAR v2.2, developed by the US EPA, was employed to predict the toxicity of the intermediates and final degradation products.

2.3. Antibacterial activity evaluation by disk diffusion method

The antibacterial activity of DOX solutions sampled at different time intervals during photocatalytic degradation using electrochemically synthesized TiO₂ was assessed via the disk diffusion method, following CLSI guidelines (CLSI, 2020). Gram-negative Escherichia coli and Gram-positive Staphylococcus aureus were used as test organisms. Bacterial suspensions were standardized to 0.5 McFarland (10⁶ CFU/mL) and spread uniformly on Tryptone Soya Agar (TSA) plates. Four wells (6 mm diameter) were created on each plate, filled with 50 µL of sample, and incubated at room temperature for 1 h to allow diffusion, followed by incubation at 37°C for 24-48 h. Inhibition zones were measured using a digital calliper to evaluate the residual antibacterial activity of the samples.

3.1. Preliminary test and material characterization

To select appropriate calcination temperatures, a preliminary test for the effect of material calcination temperature on its photocatalytic activity was conducted. As shown in Fig. S3, the calcination temperature significantly impacts the degradation efficiency of DOX. The photocatalyst calcined at 300^°C demonstrated the highest efficiency, starting at 68% and reaching 86% after 120 min, indicating its strong and consistent activity. At 400^°C, the efficiency slightly decreased and peaked at 83%, and a similar trend was observed at 500^°C with efficiency reaching 82%. However, as the calcination temperature increased to 600^°C, a more noticeable drop in efficiency was observed, starting at 57% and reaching 76% after 120 min. The catalyst calcined at 700^°C showed the lowest performance, with efficiency beginning at 50% and reaching only 64% after 120 min. These results highlight that lower calcination temperatures (300^°C) are appropriate for maintaining high photocatalytic activity, while higher temperatures (600^°C and above) may lead to reduced efficiency due to potential structural changes in the TiO₂. Moreover, when compared with the commercial TiO₂ (P25), which achieved only 71% degradation efficiency after 120 min under the same conditions, the synthesized material calcined at 300°C still demonstrates superior photocatalytic performance. Hence, a suitable calcination temperature of 300^°C (T@300) was selected for further experiments.

As can be observed from Fig. 1(a), the SEM image of TiO₂ synthesized via the electrochemical process and subsequently calcined at 300^°C reveals a granular morphology with particle sizes in the range from nanometers to sub-micrometers. The particles appear to be uniformly distributed with a rough surface texture, which is characteristic of materials synthesized at relatively low calcination temperatures. This morphology suggests the presence of a high surface area, which should be beneficial for photocatalytic applications as it provides more active sites for the degradation of contaminants like DOX. Fig. 1(a) also reveals particle agglomeration, suggesting inter-particle interactions, and the underlying cause is likely influenced by the controlled calcination conditions.

Close

Fig. 1.

SEM images of (a) T@300 (catalyst calcinated at 300^°C), (b) EDX spectrum of T@300, and elemental mapping of (c) Ti and (d) O.

Export to PPT

In Fig. 1(b), the EDX analysis of TiO₂ reveals key insights into its elemental composition and potential impurities. It confirms the successful formation of TiO₂ by the electrochemical process with the significant presence of Ti and O peaks as the dominant elements in the synthesized TiO₂. Figs. 1 (c and d) indicate a consistent distribution of titanium and oxygen elements across the sample, which suggests the synthesis method successfully generated a homogeneous material. The identification of Cl and K peaks suggests the presence of residual KCl from the electrolyte during synthesis. Further refinement of the synthesis or post-synthesis purification steps may be necessary to reduce these impurities, especially for applications requiring high purity. When combined with XRD results in Fig. 2(a), which confirm the anatase crystalline phase, the EDX data provides a robust confirmation of both the elemental and structural integrity of the TiO₂. This ensures that the material would be suitable for its intended applications, such as photocatalysis, where purity and correct composition are crucial.

Close

Fig. 2.

(a) XRD pattern and (b) FTIR spectrum of T@300.

Export to PPT

The XRD analysis revealed several peaks corresponding to both anatase and rutile phases. Peaks at 25.254^° (101) and 47.695^° (200) indicate the presence of well-crystallized anatase, a phase known for its high photocatalytic activity (Fig. 2a). The peaks at 28.356^° (110), 40.519^° (200), and 57.385^° (220) suggest a significant rutile phase, which likely contributes to enhanced electron-hole separation when combined with the anatase phase, thereby improving photocatalytic efficiency. The presence of multiple rutile peaks, such as 53.817^° (211) and 62.818^° (002), indicates that the KCl electrolyte may have facilitated the formation of rutile at lower temperatures. KCl could act as a mineralizer, promoting the nucleation and growth of rutile crystals within the TiO₂ matrix. Additionally, residual KCl might influence the electronic properties and catalytic performance of the material.

In Fig. 2(b), the FTIR spectrum of TiO₂ provides crucial insights into its structural and chemical properties. Key features to analyze are the absorption bands associated with Ti–O bonds, particularly the stretching vibrations found around 700-800 cm^-1. This range indicates the presence of Ti–O–Ti bonds, with the exact position revealing the phase of TiO₂. Anatase typically shows peaks around 700 cm^-1, while rutile shifts to slightly higher wavenumbers. The low intensity of the peaks in the Ti–O stretching region could indicate a low degree of crystallinity or incomplete phase transformation. For well-crystallized TiO₂, one expects more pronounced and sharper peaks corresponding to Ti–O–Ti bonds. Low peak intensity might reflect an amorphous or poorly crystalline material where the Ti–O bonds are less well-defined, leading to less distinct FTIR signals. Another aspect to consider is the presence of any additional peaks or shifts in the spectrum. Peaks at higher wavenumbers, such as those around 1000 - 1200 cm^-1, might indicate the presence of hydroxyl groups or adsorbed water, which is common in TiO₂ samples synthesized under aqueous conditions. The calcination process at 300^°C is expected to remove a significant amount of these hydroxyl groups, resulting in a decrease or absence of these peaks in the FTIR spectrum. Alternatively, the low intensity could also result from a lower concentration of TiO₂ in the sample or suboptimal experimental conditions. In either case, these observations suggest that the material might not be fully optimized for applications requiring high crystallinity and well-defined phase composition.

In the context of TiO₂ synthesis using KCl as an electrolyte, potassium ions likely become incorporated into the TiO₂ structure during the electrochemical process, followed by high-temperature calcination. Similar to the molten salt technique used to grow hollandite crystals (K_x(M,Ti)₈O₁₆) (Moetakef et al., 2014), KCl acts both as a solvent and a source of potassium in the TiO₂ synthesis. With increasing temperature, potassium ions can gradually diffuse into the TiO₂ lattice. Molten salts including KCl and KNO₃, under high-temperature conditions, interact with Ti and other transition metals, critically influencing crystal formation. In this study, the presence of KCl may influence the crystallization process by promoting phase growth and transformation. Its effect likely supports the development of rutile from precursor phases. Because increasing the calcination temperature likely promotes potassium incorporation into the lattice, lower calcination temperatures could limit this incorporation, thereby suppressing rutile-phase development. This is observed in TiO₂ synthesized at lower calcination temperature as indicated by XRD analysis in Fig. 2(a). Thus, the role of potassium highlights the influence of electrochemical synthesis conditions, particularly calcination temperature and KCl composition. The coexistence of anatase and rutile phases is significant for photocatalytic applications. While anatase generally exhibits higher activity due to its high surface area, bandgap, and surface defects, rutile can enhance charge separation, thus reducing the recombination of electrons and holes. The structural and electronic effects of residual KCl within the TiO₂ matrix could further influence its photocatalytic performance, particularly in the degradation of antibiotics.

An interesting discovery in the electrochemical synthesis of TiO₂ was made during the anodization of a titanium (Ti) rod (0.5 × 15 cm) in a KCl aqueous solution, and the synthesis mechanism was proposed and has been demonstrated in Fig. 3. During the electrochemical synthesis of TiO₂, particle aggregation occurs as the nanoparticles naturally tend to minimize surface energy. As TiO₂ particles nucleate and grow, their surface charge dissipates or becomes unbalanced, reducing electrostatic repulsion and promoting aggregation (Reetz et al., 1995). This clustering behavior may resemble the effect of surfactants, where particles appear bound together. Ionic species, such as K⁺ and Cl^- ions in a KCl solution, significantly influence aggregation. These ions adsorb onto the TiO₂ surface, altering the zeta potential and reducing repulsive forces between particles, which can promote coagulation, especially at higher ionic strengths. In contrast, pure water lacks sufficient ions to stabilize the surface charge, leading to accelerated aggregation. Electrolyte bridging, where adsorbed ions such as K⁺ or Cl^- act as a bridge between particles, is another mechanism contributing to coagulation. This, combined with hydrogen bonding and defect-related interactions, results in cohesive particle networks resembling surfactant-induced binding (Reetz and Helbig, 1994). Additionally, adsorbed by-products or organic impurities can form pseudo-surfactant layers on TiO₂ particles, providing steric stabilization by reducing direct particle contact. However, these layers may be unstable and promote coagulation by creating “sticky” points that facilitate particle binding. Surface hydroxyl groups (–OH) and oxygen vacancies or defects on TiO₂ particles can further promote aggregation through hydrogen bonding and defect-site interactions.

Close

Fig. 3.

Schematic diagram of electrochemical synthesis for the formation of TiO₂.

Export to PPT

3.2. Photocatalytic degradation of DOX antibiotic using T@300 as photocatalyst

The results of the antibiotic degradation efficiency using TiO₂ synthesized by an electrochemical process and subsequently calcinated at 300^°C reveal a clear influence of pH on the degradation process. As can be seen in Fig. 4(a), the degradation efficiency varied significantly across different pH levels, with distinct patterns emerging. Under acidic environments (pH 3 and 5), the degradation efficiency was relatively high, showing consistent improvement over time. At pH 3, the photocatalytic efficiency exhibited a notable enhancement, rising from approximately 66.8% at the initial stage (after 30 min of dark-phase stirring, DPS) to 80.7% after 60 min. A similar trend was observed at pH 5, where efficiency increased from 67.1% to 84.0% within the same duration. Thus TiO₂ demonstrates superior photocatalytic performance under acidic conditions, and the likely mechanism involves the positively charged TiO₂ surface, which enhances antibiotic adsorption and accelerates the photocatalytic degradation efficiency. Among all tested pH levels, the neutral condition (pH 7) yielded the highest degradation efficiency. Initially, the efficiency was recorded at 67.7%, but it exhibited a substantial rise, reaching 85.9% within 60 min. This result suggests that TiO₂ functions optimally under neutral pH, likely due to a well-balanced interaction between the catalyst surface and antibiotic molecules. However, under alkaline conditions (pH 9 to 11), the degradation efficiency exhibited a distinct hindering effect. At pH 9, the initial efficiency was recorded at 50.3%, with only a marginal increase to 52.5% after 60 min. Similarly, at pH 11, degradation remained consistently hovering approximately around an average of 50%, starting at 49.2% and peaking at merely 51.6% by the end of the experiment. The mostly unchanged tendencies in degradation efficiency performance under alkaline conditions are likely due to repulsion between the negatively charged catalyst surface and anionic antibiotic species, thereby hindering the adsorption and subsequent photocatalytic degradation. These findings emphasize the crucial role of solution pH in enabling the photocatalytic degradation of DOX by TiO₂. Notably, acidic and neutral conditions enhance the degradation process, with neutral pH offering the most suitable level of antibiotic degradation efficiency. On the other hand, alkaline conditions inhibit the degradation process, likely due to adverse surface interactions and diminished production of ROSs.

Close

Fig. 4.

Effect of (a) solution pH and (b) catalyst-to-antibiotic ratio on the photocatalytic degradation efficiency of DOX using T@300.

Export to PPT

Fig. 4(b) depicts the influence of the catalyst-to-antibiotic ratio on the DOX degradation efficiency. There was a distinct improvement in efficiency with increasing catalyst-to-antibiotic ratio (or TiO₂ concentration) from R1:5 (50 mg/L) to R1:20 (200 mg/L) when the initial DOX concentration remained unchanged. Notably, the degradation efficiency of the experiment with a ratio of R1:5 initially began at approximately 51% and reached only 61.76% after 60 min. However, using a ratio of R1:20 progressively reached 85.88% after 60 min. This result suggests that increasing the catalyst-to-antibiotic ratio or initial concentration of TiO₂ could provide additional active sites for the photocatalytic reaction, thereby more effective degradation of DOX. However, exceeding the R1:20 value leads to a marginal decline in efficiency. Specifically, as the catalyst concentration increases from R1:30 to R1:50, a decrease trend becomes apparent. This reduction in degradation efficiency using higher catalyst concentration may be attributed to the agglomeration of TiO₂ particles, which limits the accessible surface area for catalytic activity. Additionally, excessive catalyst loading can intensify light scattering, thereby preventing absorption and hindering photocatalytic performance. While an initial increase in catalyst dosage enhances degradation, surpassing a critical threshold leads to diminished efficiency. The ratio of R1:20 (200 mg/L of catalyst) exhibits superior photocatalytic performance in DOX degradation, which highlights the critical importance of precisely regulating catalyst dosage to enhance degradation efficiency.

The influence of initial antibiotic concentration on degradation efficiency has been depicted in Fig. 5(a). At lower DOX concentrations, particularly at 5 and 10 mg/L, the TiO₂ photocatalyst exhibited notable degradation performance, and a significant fraction of the antibiotic underwent degradation within a short timeframe. For instance, at 5 mg/L, the degradation efficiency initially reached ∼70.65% and further rose to 75.47% after 60 min, indicating that the catalyst is capable of breaking down DOX even at low concentrations. Despite its capability, at lower concentrations (e.g., 5 mg/L), the photocatalytic process may be less effective due to the limited presence of antibiotic molecules, thereby restricting adsorption onto the catalyst surface and subsequently slowing degradation kinetics. In contrast, higher antibiotic levels enhance catalyst-molecule interactions, yielding a more distinct kinetic profile and improving the reliability of degradation assessments. Therefore, excluding concentrations lower than 10 mg/L from the kinetic analysis not only improves the reliability of the data but also ensures that the study results are more applicable to actual environmental scenarios where higher concentrations of antibiotic residues are more likely to be encountered.

Close

Fig. 5.

Kinetic analysis of DOX photocatalytic degradation: (a) DOX degradation at different concentrations using TiO₂ photocatalysis, (b) Langmuir-Hinshelwood model fit for DOX degradation at varying concentrations and corresponding rate constants.

Export to PPT

As the initial DOX concentration increased to 15 and 20 mg/L, the degradation efficiency decreased. Although the catalyst still performed adequately, the efficiency started to plateau. For DOX concentrations of 15 mg/L, the efficiency rose from 70.55% to 86.13%, while for 20 mg/L and 30 mg/L, it respectively reached 81.62% and 76.86% after 60 min, indicating that at high concentrations, the photocatalytic activity is substantially hindered. This could be due to the saturation of the active sites on the catalyst surface or the increased optical density of the solution, which limits the penetration of UV light necessary for the activation of the TiO₂ photocatalyst. Overall, these results highlight that while T@300 is effective for degrading DOX at lower concentrations, its efficiency decreases as the initial antibiotic concentration increases, pointing to potential limitations in its application for treating higher concentrations of pollutants in water.

The kinetic analysis of DOX photocatalytic degradation using TiO₂ would reveal important insights into the degradation mechanism. As can be seen from Fig. 5(b), distinct patterns emerged across different initial concentrations of DOX (10, 15, 20, and 30 mg/L). As previously aforementioned, a low concentration of 5 mg/L was excluded from the kinetic analysis as evidenced by significantly lower degradation rates (k_obs = 0.0192 min^-1) and poor correlation with the first-order kinetic model (R² = 0.6303) in Fig. 5(b). These lower concentrations may not reflect real-world scenarios, where antibiotic residues in polluted aquatic environments are typically higher, often exceeding 10 mg/L due to sources like pharmaceutical wastewater and agricultural runoff (Xu et al., 2021). Thus, other tested concentrations expressed with the coefficient of determination more well-fitted (DX10 with R² = 0.99, DX15 with R² = 0.9789, DX20 with R² = 0.964, and DX30 with R² = 0.995) were chosen for later discussions. It was observed that the higher the initial concentration of DOX introduced at the beginning of the experiment, the slower the degradation rate. This phenomenon is consistent with the Langmuir-Hinshelwood (L-H) model, which describes the relationship between adsorption and surface-mediated reactions. In addition, the system contains more DOX molecules than the capability of TiO₂ to adsorb on the surface, thereby leading to difficulty in maintaining high degradation rates. This reduction in efficiency suggests potential limitations in the reaction mechanism (e.g., lack of ROSs for further degradation) or external influences (e.g., scattering light effect). Specifically, at an initial concentration of 10 mg/L, ln(C₀/Ct) rose from 1.13 to 1.85 after 60 min. Similar trends emerged at other concentrations, though with varying degradation rates. Moreover, the observed rate constant (k_obs), derived from the slope of the first-order kinetic model, exhibited a declining trend as the initial DOX concentration increases (as seen in Fig. 5a). Specifically, the k_obs values were 0.1006, 0.0899, 0.0703, and 0.0441 min^-1 for DOX concentrations of 10, 15, 20, and 30 mg/L, respectively.

As depicted in Fig. 5(b), the L-H model provides additional validation. The derived parameters, including the surface reaction rate constant (K_C) and adsorption equilibrium constant (K_L-H), reveal the dynamic interaction between adsorption processes and catalytic surface reactions. The model suggests that at higher concentrations, the adsorption sites on the TiO₂ surface become saturated more quickly, leading to a reduction in the overall degradation rate. In summary, the kinetic data suggest that DOX degradation using TiO₂ follows a pseudo-first-order reaction at lower concentrations and transitions to L-H kinetics as the concentration increases. The decreasing k_obs with higher concentrations highlight the importance of catalyst loading and concentration control in optimizing the photocatalytic degradation process. These results provide valuable insights for improving the efficiency of TiO₂-based photocatalytic systems for antibiotic degradation in water treatment applications.

3.3. Radical scavenging test and photocatalytic mechanism

The inset of Fig. 6 demonstrates the effect of various radical scavengers (i.e., p-BQ, AO, NaN₃, and TBA at a concentration of 1 mM). This radical scavenging test provides crucial insights into the contributions of various ROSs to the photocatalytic degradation of DOX. In the absence of any scavenger, the degradation efficiency reached a maximum of 85.88% after 60 min. However, the addition of selective scavengers induced notable variations in efficiency, highlighting the contributions of various ROSs. The introduction of NaN, a known quencher of ¹O₂ and ^•OH, led to a marginal reduction in degradation efficiency, bringing it down to 82.17%. This slight decline implies that these ROSs are not the key oxidants of the reaction. Similarly, the application of DiA, an established scavenger of photogenerated holes (h⁺), resulted in a further but still modest decrease in efficiency to 79.39%. The small reduction in degradation performance suggests that h⁺ also does not play a significant role in the process. Hence, the reaction mainly followed alternative mechanisms or rather different ROSs than ¹O₂, ^•OH, and h⁺.

Close

Fig. 6.

Proposed degradation pathways of DOX in the photocatalytic reaction by TiO₂.

Export to PPT

In other scavenger trapping experiments, it was found that the most substantial drop in degradation efficiency occurred when pBZ was introduced as a superoxide radical (^•O₂^-) scavenger. The addition of pBZ caused the degradation efficiency to hover around an average of 70%, with only slight fluctuations from 68.63% up to 71.97%, indicating that a considerable suppressing effect took place. This significant hindering by the introduction of pBZ highlights the critical role of ^•O₂^- in the degradation mechanism. It took notice that the efficiency rates slightly declined from 71.97% (at 25 mins) to 68.63% (at 60 mins). Because the degradation process was prevented, DOX adsorbed on the TiO₂ surface underwent the desorption phase. This also happens in the experiment with the addition of TeB, known as a hydroxyl radical (^•OH) scavenger. TeB exhibited a modest variation around an average of 72% without significant deviation over 60 min. It showed a regular upper-efficiency rate of 74.38% and a lower rate of 68.81%. The addition of TeB, hindering the degradation from occurring, indicates that ^•OH radicals participate in the degradation process. However, when compared to pBZ, TeB showed a lesser influence in the hindering effect, thereby ^•OH radicals are to some extent less dominant compared to ^•O₂^-. These findings highlight that the decomposition of DOX by TiO₂ is facilitated by a complex mechanism involving multiple ROSs. The relative contributions of the ROSs can be exhibited by the following trend. ^•O₂^- exhibited the dominant influence, followed by ^•OH with slightly less impact. While h⁺ contributed to some extent, ¹O₂ had the least impact among ROSs.

Based on the understanding of ROS generation and their activity, the photocatalytic mechanism was proposed. The photocatalytic degradation of DOX using TiO₂ follows a sequence of redox reactions. These transformations are triggered upon exposure to UV irradiation, initiating a series of oxidative processes. The process begins with the photoexcitation of TiO₂, where photons with energy equal to or greater than the band gap excite electrons from the valence band (VB) to the conduction band (CB), creating electron-hole pairs (Reaction 5). The photogenerated electrons (e^-) in the CB reduce molecular oxygen (O₂) adsorbed on the TiO₂ surface, forming ^•O₂^- (Reaction 6). ^•O₂^- plays a significant role in degrading DOX either directly or by reacting with protons to produce H₂O₂ (Reaction 7). H₂O₂ further decomposes under UV light to form highly reactive ^•OH (Reaction 4). Simultaneously, the h⁺ in the VB oxidizes H₂O or OH^- to generate ^•OH (Reactions 9 and 10). In addition to ^•O₂^- and ^•OH, ¹O₂ is produced, likely through energy transfer processes or secondary reactions involving ROSs like ^•O₂^-. Together, these ROSs interact with DOX, leading to its breakdown into smaller, less harmful byproducts (Reaction 11). By understanding the contributions of these reactions and their interdependencies, the photocatalytic system can be optimized to enhance ROS production, improving the degradation efficiency of pollutants like DOX in environmental applications. The complete reaction mechanism illustrating the generation and roles of ROSs in the photocatalytic degradation of DOX has been summarized in Fig. 6, highlighting the interconnected mechanisms and key reactions involved in the process.

3.4. DOX degradation pathways and toxicity predictions

The degradation of DOX under UV irradiation was analyzed through the progression of fragment ion intensities in the liquid chromatography tandem mass spectrometry (LC-MS/MS) spectra (Fig. S4). Intermediates identified during DOX degradation suggest hydroxylation and ring cleavage processes (see Table S3 for details on identified compounds and their m/z values). Initially, spectra recorded after 30 min of mixing without UV exposure (DPS phase, see Fig. S4b) indicated minimal changes in the intensities of ions with higher mass-to-charge ratios (m/z), including the parent DOX ion at m/z = 444 compared to Fig. S4(a). This result suggests that the catalyst remained inactive without UV energy, and no significant degradation occurred during the adsorption equilibrium phase. Interestingly, an additional ion at m/z = 445, attributed to the isomerization of the parent molecule, was also detected, indicating potential structural rearrangements in DOX even before UV exposure (Zhong et al., 2022). Upon UV irradiation, significant transformations in the fragment ion intensities were observed. By 5 min of exposure (Fig. S4c), the intensity of the parent ion (m/z = 444) began to decrease, marking the initiation of the degradation process. The decline in parent ion intensity is attributed to bond cleavage facilitated by the activated catalyst and UV energy. By 10 min (Fig. S4d), several degradation products emerged, including prominent fragments at m/z = 361, 158, and 181. These fragments are associated with pathways involving the cleavage of amide bonds, ester groups, or hydroxyl-containing side chains, along with structural rearrangements and loss of functional groups (Chen et al., 2024; Shen et al., 2024). After 15 min (Fig. S4e), the major ion at m/z = 361 further fragmented into smaller products with m/z = 343 and 304. These transitions involved the stepwise loss of functional groups, such as hydroxyl (–OH), carboxyl (–COOH), or carbonyl (C=O), as well as molecular rearrangements. The continuous reduction in intensity of these intermediates highlighted the progression of advanced degradation. By 30 min, most high-mass intermediates had disappeared, indicating the effective breakdown of DOX under the photocatalytic process. At 120 min (Fig. S4f), the dominant ions observed included m/z = 57, 82, and 214. These low-mass fragments represent stable byproducts formed through successive bond cleavages and oxidation processes. The ion at m/z = 57 likely originated from alkyl chain cleavage or the removal of functional groups during hydrolysis, while m/z = 82 may result from heterocyclic ring cleavage or the loss of hydroxyl- or amide-containing groups.

The degradation of DOX under UV irradiation is proposed with three main pathways (Fig. 7). In the first pathway, the parent DOX ion (m/z = 444) undergoes demethylation and deamidation to produce DOX1 (m/z = 416) and DOX2 (m/z = 400), respectively (Chen et al., 2024; Shen et al., 2024). These intermediates are further oxidized and fragmented, leading to products such as DOX3 (m/z = 356), DOX4 (m/z = 343), and DOX5 (m/z = 304). These transformations involve the loss of methylamine, hydroxyl groups, and ring-opening reactions, resulting in a stepwise breakdown (Zhong et al., 2022; Chen et al., 2024). The second pathway primarily involves the detachment of N-methyl and amino groups, as well as the opening of benzene rings. DOX first undergoes demethylation at C6 and dedimethylation at C14, forming DOX6 (m/z = 430) (Hong et al., 2020). Under oxidative stress, DOX6 reacts with ^•O₂^- or h⁺ to produce DOX7 (m/z = 362). Subsequent delamination, dehydroxylation, and decarbonylation at positions C11, C8, C7, and C10 result in the formation of DOX8 (m/z = 184) (Xiong et al., 2018; Yan et al., 2021).

Close

Fig. 7.

Proposed mechanism for the photocatalytic reaction and degradation of DOX using electrochemically synthesized TiO₂ (inset: scavenging test to determine the formation and role of ROSs in the DOX degradation).

Export to PPT

In pathways IIIA and IIIB, the ion fragment at m/z = 445, formed via isomerization of DOX (m/z = 444) (Zhong et al., 2022), is a monohydroxylation byproduct. The C11a–C12 double bond is more susceptible to ROS attack than C2–C3, due to adjacent electron-donating and electron-withdrawing groups, making it the primary hydroxylation site (Zhong et al., 2022). Fragmentation of these hydroxylation products, including the loss of NH₃, aligns with previously reported mechanisms (Zhong et al., 2022). However, DOX resists rearrangement at C12, likely due to steric and electronic effects from the hydroxyl group at C5. Hydrolysis of DOX (m/z = 444) produces the fragment ion at m/z = 214 through cleavage of key functional groups, such as amino and lactone groups (Wu et al., 2022). Further oxidation results in smaller degradation products (DP1, DP2, and DP3) with m/z = 82, 74, and 57, illustrating the breakdown of DOX into simpler intermediates under experimental conditions (Yan et al., 2021).

The potential ecological toxicity of DOX and its degradation products, including DOX7, DOX10, and DP1 to DP3, was predicted using the ECOSAR software (v2.2) across three aquatic organisms: fish, daphnids, and green algae (Table 1). For DOX, the results indicate that while it remains relatively harmless to fish and daphnids, it is highly toxic to green algae. This is reflected in its low effective concentration (EC50) value for algae (3 mg/L), indicating significant toxicity at low concentrations. DOX also shows toxicity in chronic exposure conditions for both fish and daphnids, with chronic toxicity (ChV) values of 45 and 8 mg/L, respectively. Despite its toxicity to green algae, DOX is classified as harmless under both acute and chronic conditions for other aquatic species. In contrast, the degradation by-product DOX7 exhibits the highest lethal concentration (LC50) and ChV values among selected compounds, considered harmless to fish, daphnids, and algae. Therefore, DOX7 is unlikely to pose a significant ecological risk. Similarly, DOX10 also exhibits minimal toxicity, with its LC50 and ChV values indicating its classification as harmless to the same organisms. This further supports the belief that the breakdown of DOX yields degradation intermediates with notably reduced toxicity.

However, the intermediate derivatives DP1, DP2, and DP3 exhibit distinct toxicity characteristics. Notably, DP1 emerges as a highly toxic by-product, exhibiting distinct toxicity across all tested organisms. Its acute toxicity is apparent from LC50 values as low as 32 mg/L for fish and 23 mg/L for daphnids. Additionally, DP1 demonstrates notable toxicity toward green algae, with an EC50 of 58 mg/L. This indicates that DP1, as an intermediate, may present substantial environmental hazards if persistent accumulation occurs. In contrast, DP3 presents negligible toxicity, as indicated by its elevated LC50 and ChV values across all tested organisms. However, DP2 presents a more complex toxicity profile. Although DP2 presents values of LC50, ChV, and EC50 lower than those of the parent antibiotic, DP2 may still pose some ecological concern, particularly in long-term exposure scenarios, despite being less harmful than DP1. The degradation of DOX predominantly yields transformation products with reduced or negligible toxicity. However, an exception is observed in DP1, which exhibits noticeable toxicity across diverse biological species. This finding highlights the necessity of analyzing intermediate degradation by-products in environmental conditions. Certain derivatives, such as DP1, may present a greater ecological hazard than the parent compound, demanding cautious assessment. However, degradation products like DOX7, DOX10, DP2, and DP3 are likely to exhibit minimal toxicity toward aquatic organisms. This suggests that the ultimate breakdown of DOX could predominantly generate compounds with insignificant environmental risk.

In the antibacterial activity, Figs. S5 and S6 show the results in inhibiting the growth of Staphylococcus aureus and Escherichia coli using the disk diffusion method. The untreated DOX solution (DOX-TV) exhibited strong antibacterial activity against both bacteria, reflecting its effective functional groups for antimicrobial activity. However, at 5 min (5TV), E. coli showed complete resistance (0 mm), but S. aureus still exhibited inhibition zones (11.07 ± 0.48 mm), indicating that initial degradation fragments had already altered the structure-function relationship of DOX, particularly in ways that impair its action on Gram-negative organisms. This might be an indication of bond cleavage involving demethylation and deamidation, initiating the breakdown of pharmacologically active sites. Simultaneously, the formation of initial degradation products such as DOX1 (m/z = 416) and DOX2 (m/z = 400) may retain partial bioactivity, particularly against S. aureus. However, the antibacterial properties of DOX against E. coli were completely lost after 5 min, as the parent compound broke down into inactive forms that are evidently ineffective.

As degradation progressed further, the inhibition zones against S. aureus persisted (14.81 ± 0.52 mm at 10 min, 13.92 ± 0.45 mm at 15 min, and 12.51 ± 0.67 mm at 30 min). According to the proposed degradation pathways in Fig. 7 and MS spectra in Fig. S4, this period corresponds to the emergence of deeper structural modifications such as DOX3 (m/z = 356), DOX4 (m/z = 343), DOX5 (m/z = 304), and the predominance of smaller ions such as m/z = 214, 158, and 181. These intermediates likely possess antimicrobial effects on Gram-positive bacteria. Notably, by 45 min, all tested samples (45TV, 60TV, 90TV, and 120TV) exhibited no antibacterial activity (0 mm inhibition zones), strongly indicating complete degradation of DOX or transformation into inactive products in antimicrobial activities.

The loss of antibacterial activity is also in line with the toxicity predictions performed using ECOSAR software. While the parent DOX molecule was found to be highly toxic to green algae and moderately harmful under chronic exposure to fish and daphnids, its degradation products, particularly DOX7 and DOX10, were classified as harmless across all tested aquatic organisms. However, caution is warranted due to the predicted toxicity of intermediate product DP1, which displayed significant acute and chronic toxicity across fish, daphnids, and algae. While this compound may not have persisted at biologically significant concentrations in the final stages of degradation (as suggested by its absence in the disk diffusion results at 45 - 120 min), its potential formation highlights the importance of complete mineralization in photocatalytic systems to avoid ecological risks from transient toxic by-products. However, this study does not include experimental cytotoxicity evaluations of the electrochemically synthesized TiO₂ nanoparticles, as such assessments require in vitro or in vivo models beyond the material performance and environmental focus of this work.