Photodegradation of tetracycline hydrochloride by biochar/Cu₂O coupled with persulfate: Insights into the factors and intermediates toxicity

2.1. Chemical reagents and catalyst preparation

All chemicals were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (China) with the following specifications: Chitosan: Deacetylation degree = 80-95%. Acetonitrile (for HPLC mobile phase): HPLC grade, ≥99.9% purity. Other reagents (CuSO₄, NaOH, glucose, PS, etc.): Analytical reagent grade, ≥99.0% purity. Ultra-pure water (18.2 MΩ·cm resistivity) was prepared using a Millipore Milli-Q system.

BC was prepared using chitosan as biomass according to our previous study with some modifications [29]. The preparation process was as follows: Chitosan was pyrolyzed at 600°C (5°C/min) under N₂ flow and held for 2 h. The resulting material was washed with 0.1 M HCl and deionized water until neutrality (pH ≈ 7.0). The neutralized product was then pyrolyzed at 800°C (5°C/min, N₂ flow) for 2 h. The final BC was designated BC-800.

The Cu₂O nanoparticles were synthesized using the chemical precipitation method and loaded on the BC-800. The specific process was as follows: 1.25 g CuSO₄ was dissolved in 200 mL of water, then 0.8 g NaOH was added and stirred to obtain Cu(OH)₂ precipitation. Next, 100 mL of 1 mg/mL BC-800 suspension was added to the Cu(OH)₂ suspension and stirred at 60°C. After that, 1 g glucose (used as a reductant) was added to the suspension and reacted for 1 h, the Cu(OH)₂ was reduced to Cu₂O and loaded on the BC-800. The resulting dark brown precipitate (designated BC/Cu₂O) was washed thoroughly with ethanol and deionized water, followed by drying at 55°C for 24 h in a vacuum oven.​

2.2. Catalyst characterization

Scanning electron microscope (SEM, JEOL, JMS-IT800is) equipped with energy dispersive spectroscopy (EDS) was employed to observe the morphological structure of BC/Cu₂O. Brunauer-Emmett-Teller (BET) surface areas and pore size distribution analyses were performed using a Micromeritics ASAP 2460 instrument. The chemical composition was determined by X-ray diffraction (XRD, Bruker D8 instrument, Cu-Kα radiation) and X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific Escalab 250Xi). Electron Paramagnetic Resonance (EPR, JEOL FA200) was used to detect reactive oxygen species during degradation. Sulfate radical (SO₄^•⁻) and hydroxyl radical (·OH) were trapped in a water solution, ·O2 was detected in an ethanol system, and the hole was examined using the BC/Cu₂O powder. Electrochemical impedance spectroscopy (EIS) measurements followed our previously reported method [29].

2.3. Batch experiments

A typical degradation experiment ​​was conducted​​ in a 250 mL Erlenmeyer flask containing 100 mL of 50 mg/L TC-HCl and 0.005 g BC/Cu₂O. The flask was placed on a magnetic stirrer with a stirring speed of 250 rpm and room temperature (28°C). A CEL-HXF300 Photocatalytic Xenon Lamp​​ (>300 nm, 300W, Zhongjiao Jinyuan, Beijing, China) was used as a light source. The irradiance is 220 ± 5 mW/cm²​ (working current of 21A) measured by a photo-radiometer. And the distance between the reaction liquid level and the xenon lamp is 16 cm during the degradation process. The TC-HCl solution was collected and filtered by a 0.22 μm PTEE syringe filter and immediately analyzed via high-performance liquid chromatography (HPLC) (Waters 2695). The first sample was withdrawn and determined after 5 min of reaction. The mobile phase consisted of acetonitrile/0.1% formic acid (22/78, v/v) with a flow rate of 0.2 mL/min, with column temperature of 25°C. The pseudo-first-order kinetic model was applied to calculate the apparent rate constant (kobs​) of TC-HCl. The integrated rate law for pseudo-first-order kinetics is expressed as shown in Eq. (1):

where C₀ and C_t are the TC-HCl concentration (mg/L) at the initial time and time t (min). A linear regression of ln(C₀/C_t) vs. t was performed, where the slope corresponds to k_obs​. The total organic carbon (TOC) of the solution before and after treatment was analyzed (Shimadzu, Japan). A ​​Box-Behnken response surface methodology (RSM) design​​ optimized parameters ​​with three independent variables​​: PS concentration, pH, ​​and xenon lamp current​​. The design ​​included three levels per factor​​ with three central point replicates, ​​totaling 18 experiments​​. All experiments ​​were performed in triplicate. All batch experiments (except RSM) were conducted in triplicate, and the mean values were presented. HPLC-MS (Agilent Technologies) was applied to identify the degradation intermediates of TC-HCl.

3.1. Characterization analysis

As shown in Figure 1(a), BC-800 exhibited a macroporous structure (100-500 nm pores). After Cu₂O loading (Figure 1b), cubic nanoparticles were anchored on BC-800, attributed to the coordination between BC surface groups and Cu²⁺ ions during chemical co-precipitation. XRD patterns (Figure 1c) showed six characteristic peaks at 29.6°, 36.5°, 42.4°, 61.5°, 73.6°, and 77.7°, matching Cu₂O (JCPDS 05-0667) with no detectable CuO impurities. Elemental mapping (Figures 1d-g) of the region in Figure 1(b) confirmed homogeneous distributions of Cu, O, C, and N, with XPS revealing atomic percentages of 7.24%, 12.55%, 77.51%, and 2.70%, respectively. BET analysis (Figure S1) further demonstrated a mesoporous structure (average pore size: 3.07 nm; surface area: 234 m²/g), suggesting hierarchical porosity beneficial for reactant diffusion.

Close

Figure 1.

SEM images of (a) BC-800 and (b) BC/Cu₂O; (c) XRD pattern of BC/Cu₂O and standard Cu₂O; elements mapping images of (d) C, (e) O, (f) Cu, (g) N.

Export to PPT

3.2. Performances of photodegradation of TC-HCl by different catalysts

Adsorption experiments showed that single BC/Cu₂O could only remove 30% of TC-HCl. BC/Cu₂O as a catalyst exhibited significantly enhanced TC-HCl degradation under visible light with PS, achieving 98% removal within 90 min (Figure 2a). This performance surpassed control systems of bare Cu₂O/PS (85% removal), BC/Cu₂O without PS (63% removal), and dark BC/Cu₂O/PS (48% removal). The observed first-order rate constant (K_obs) increased from 0.0086 min^-1 (BC/Cu₂O) to 0.0355 min^-1(BC/Cu₂O/PS), representing a 4.12-fold enhancement (Figure 2b). The result was superior to TiO₂/g-C₃N₄, S-doped g-C₃N₄/BC/PS/visible light system and Cu₂O -metal organic frameworks based photocatalysis system for TC-HCl degradation [27,30,31]. The synergy factor (SF), calculated as shown in Eq. (2):

Close

Figure 2.

(a) Degradation efficiency and (b) Kinetics constant of TC-HCl by different oxidation systems, (c) EIS results of BC/Cu2O and Cu2O.

Export to PPT

confirms true synergistic interaction beyond additive effects. The enhancement was attributed to several reasons. Firstly, SEM (Figure 1b) revealed uniform dispersion of cubic Cu₂O nanoparticles on porous BC, reducing aggregation and increasing accessible active sites versus bulk Cu₂O. Secondly, EIS Nyquist plots (Figure 2c) showed a depressed semicircle for BC/Cu₂O, significantly lower than pure Cu₂O. This validates BC’s role as an electron shuttle suppressing recombination. Thirdly, PS scavenged photogenerated electrons via a thermodynamically favorable reaction (3):

where the conduction band of Cu₂O (-1.1 eV) provides sufficient driving force. This dual function includes depleting electrons to reduce electron-hole recombination and generating additional SO₄^•⁻ radicals (validated by EPR in Section 3.3). Control experiments confirmed minimal PS activation in darkness (48% removal), highlighting the light-dependent synergy. Integrating PS with BC/Cu₂O thus establishes an efficient and cost-effective advanced oxidation platform for antibiotic remediation.

3.3. Catalytic mechanism

Quenching experiments and EPR detection were used to reveal the leading reactive species in the PS-coupled photodegradation system. Methanol (MeOH, 0.25M), isopropyl alcohol (IPA, 0.25M), ascorbic acid (AA, 0.02M), and triethanolamine (TEOA, 0.02M) were used to quench SO₄^•-, ^•OH, ^•O₂, and h⁺ [32]. In considering the redox reaction between PS and AA or TEOA. Quenching of ^•O₂^- and h⁺ was conducted in a system without PS. As shown in Figure 3(a), after adding TEOA and AA, the degradation of TC-HCl decreased from 55% to 35% and 33%, and the quenching effect indicated that h⁺ and ^•O₂^- were both responsible for TC-HCl removal, and their contributions were comparable. After adding MeOH and IPA, the removal efficiency decreased from 95% to 60% and 62% (Figure 3b). It can be inferred that both SO₄^•- and ^•OH were formed in the reaction system. EPR detection also verified the existence of ^•O₂^- (Figure 3c), h⁺(Figure 3d), SO₄^•- and ^•OH (Figure 3e) in this system. Based on the quenching experiments and EPR analysis, it can be inferred that h⁺, SO₄^•-, ^•OH, and ^•O₂^- were the main radicals for TC-HCl degradation. The specific oxidation species formation processes were concluded as below:

$\text{BC}/{\text{Cu}}_{\text{2}}\text{O}+\text{light}\to {\text{h}}^{+}\left(\text{VB}\right)+{\text{e}}^{-}\left(\text{CB}\right){\text{O}}_{\text{2}}+{\text{e}}^{-}\left(\text{CB}\right){\to }^{•}{\text{O}}_{\text{2}}^{-}$

${\text{S}}_{\text{2}}{\text{O}}_{\text{8}}^{\text{2}-}+{\text{e}}^{-}\left(\text{CB}\right)\to {\text{SO}}_{\text{4}}^{•-}+{\text{SO}}_{\text{4}}^{\text{2}-}{\text{SO}}_{\text{4}}^{•-+}{\text{OH}}^{-}\to {\text{SO}}_{\text{4}}^{\text{2}-}{+}^{•}\text{OH}$

Close

Figure 3.

Quenching effect of adding (a) methanol and isopropyl alcohol, (b) ascorbic acid and (c) triethanolamine; EPR signal of •O2, (d) h+, (e) SO4•^-, and •OH.

Export to PPT

h⁺, ^•OH, and ^•O₂^- were often detected in photocatalysis systems as proved by previous studies [23,33]. The presence of SO₄^•- was due to the activation of PS by photo-induced electron.

The XPS analysis was used to evaluate the chemical valence states of BC/Cu₂O to further explore the mechanism. The O atomic percentage in BC/Cu₂O increased from 12.55% to 26.64% after usage, possibly due to the oxidation of BC/Cu₂O. High-resolution XPS spectra of C1s, O1s, and Cu2p before and after usage have been presented in Figure 4. Results showed that the C1s spectra of BC/Cu₂O contained three peaks corresponding to C=C (284.8 eV), C-O (285.7eV), and C=O (289eV) (Figure 4a); after being used as a catalyst, the C-O percentage decreased to 13.42% and the C=O percentage increased to 28.72% (Figure 4b), significantly. It is possible that some C/O sites acted as active sites during the degradation process. And for the O1s peaks, three peaks with binding energy at 529.9eV, 531.4 eV, and 533.6eV were fitted (Figure 4c), which were assigned to Cu-O, C-O, and C=O, respectively. After usage, the atomic percentage of C=O increased to 61.8% (Figure 4d), and the chemical state of Cu-O and C-O decreased, which indicated that surface oxidation of BC/Cu₂O occurred during the catalytic reaction. As shown in Figure 4(e), the binding energy of 934.3eV and 954.2eV belongs to Cu 2p_3/2 and Cu 2p_1/2 spectra, were related to Cu^+. The peak at 962.8eV was assigned to Cu 2p_1/2 and corresponded to Cu²⁺. The satellite peaks with weak energy at 941.2eV and 944.0eV were typical Cu²⁺ peaks. Besides, the peak intensity of Cu²⁺ of the used BC/Cu₂O increased due to the oxidation of Cu₂O.

Close

Figure 4.

The C1s, O1s, and Cu2p XPS spectra in (a, c, e) fresh and (b, d, f) used BC/Cu₂O.

Export to PPT

3.4. Effect of operation parameters

The effects of operational variables, including catalyst dosage, PS, and initial TC-HCl concentration, and pH on degradation efficiency were investigated. As presented in Figure 5(a), increasing the catalyst loading from 0.001 g to 0.005 g, the degradation efficiency increased from 55% to 99.1%. Further addition of catalyst to 0.015 g showed no significant improvement, likely due to the reduced light penetration at high BC/Cu₂O concentration [34]. For the effect of PS concentration (Figure 5b), increasing PS from 0.05 mM to 0.2 mM accelerates degradation; however, no obvious difference (<5%) was observed between 0.2 mM and 0.3 mM, which was attributable to radical quenching by excess PS. The degradation efficiency decreased with the increasing in TC-HCl concentration (Figure 5c), which was due to the limited active oxidative species provided by the solution system. The initial pH of 100mg/L TC-HCl was about 2.8, and the solution pH was adjusted by NaOH or HCl. The influence of pH showed that the highest catalytic activity was obtained at pH 5 (Figure 5d), while efficiency significantly declined at pH levels of 1 to 3. The structure of Cu₂O/BC might change in the strong acidic solution [25], leading to a lower degradation rate. At pH 9, the high TC-HCl degradation efficiency (93.4%) was probably due to more reactive species produced by the activation of PS under alkaline conditions [35].

Close

Figure 5.

The effect of (a) BC/Cu₂O dosage, (b) PS concentration, (c) TC-HCl concentration, and (d) initial pH on the degradation efficiency of TC-HCl by photocatalysis coupled PS system.

Export to PPT

3.5. Response surface model analysis

One-factor analysis can’t fully explore the influence of multiple factors (PS concentration, UV intensity, pH value) in this PS-coupled photocatalysis system. To investigate the key factors and their interaction, RSM was applied. The Box-Behnken model in the Design Expert software was used to design the experiments. PS concentration, pH value, and UV light intensity were chosen as the influencing factors (independent variables), and the degradation efficiency of 75 mg/L TC-HCl at the reaction time of 90 min was the response value. And the concentration of BC/Cu₂O for each treatment was set as 0.05g/L. Based on the experiment results and fitting analysis, a linear model was suggested to use. The model was R%=18.89+6.09*pH+54.34*PS+0.11*UV, with the R²=0.92. And the Analysis of Variance (ANOVA) for the linear model has been shown in Table S1. The p-value of pH and PS was less than 0.05, indicating the two parameters had a significant influence on TC-HCl degradation. And the significant of the model and not non-significance of Lack of Fit indicated that the linear model was credible. In contrast, the quadratic model showed a poor fit (R² < 0.5), likely due to the limited variation range of operational parameters in this study. Statistically, when factor ranges are narrow (<20% of the feasible domain), quadratic terms often fail to capture significant curvature while linear approximations remain robust. Consequently, the linear model was selected as it adequately represents system behavior within the tested constraints. The interaction of the various parameters has been shown in Figure 6. And optimization analysis using Design Expert showed that the best experimental parameters were pH=9, PS=0.2 mM, and UV=19.6 A, 89.6% of 75 mg/L TC-HCl could be degraded under these conditions.

Close

Figure 6.

Response surface plot presenting the interaction effect of (a) PS concentration and pH on the degradation of TC-HCl; (b) Intensity of UV light and pH on the degradation of TC-HCl.

Export to PPT

3.6. Degradation pathway of TC-HCl and toxicity evaluation

To gain deep insights into the degradation of TC-HCl in photocatalysis coupled PS system, the possible intermediates and pathways were investigated based on HPLC-MS (Figures S2 and S3) and previous literature. Besides, The Fukui functions of atoms in TC-HCl were calculated. The Fukui indices of electrophilic (f^-), radical (f⁰), and nucleophilic (f⁺) could theoretically reveal reactive sites of TC-HCl molecules [36]. The Fukui results of the main atoms have been shown in Table S1 and Figure S2. Two possible degradation pathways have been shown in Figure 7. Specifically, TC-HCl (M/Z=445) was primarily attacked by SO₄^•- or ^•OH at C8-C10 to form a hydroxylated product (B1) due to its higher electron density compared to C2-C3. This preference is supported by Fukui indices: C8(f^-=0.0244) and C10(f^-=0.198) > C2(f^-=0.004), C3(f^-=0.0103). The B2 was generated possibly through two ways: (1) the dealkylation process occurred at the C1 position and formed a hydroxyl group; (2) the double bond group of C13-C15 in B1 was attacked, forming a ketone and a carboxylic group [37]. With the further oxidation by reactive species, the double bond of C2-C3 and the hydroxyl group of C1 in B2 were attacked, and the intermediate of B3 (m/z=495) was formed. For pathway A, the double bond at C17=C18(f^-=0.0334) was most likely to be attacked by SO₄^•− to form A1 based on the prediction of oxidation position in TC-HCl by theoretical calculations [36]. Since the saturated nitrogen in C1 was oxidized by •OH, resulting in the loss of two methyl groups, and further degraded into A2 by the loss of an amide group under the attack of SO₄^•- or ^•OH [38]. C=C bond on C17=C18, C2=C3 was unstable, and C15 had a high f-value (0.0334). Carbon ring cleavage reactions occurred in A2, due to the great oxidizing performances of SO₄^•- and h⁺, leading to the formation of A3 [39]. Subsequently, by the attack of reactive oxidation species, demethylation, ring-open, and dehydroxylation reactions occurred, leading to the formation of A4 and A5. Ultimately, part of the intermediates further degraded into short-chain molecules and even mineralized into H₂O and CO₂.

Close

Figure 7.

The proposed degradation pathway of TC-HCl by the photocatalysis coupled PS system.

Export to PPT

Besides the degradation pathway, the acute and chronic toxicity of TC-HCl and its intermediates was calculated using ECOSAR to assess their environmental risk. As shown in Table 1, TC-HCl exhibited varying degrees of toxicity to fish, daphnids, and green algae, with different impacts on each. During the degradation process, although some intermediates (A2 and A3) showed higher toxicity compared with TC-HCl, the toxicity of most of the products was alleviated. After 90 min degradation, the TOC of the sample was determined. The removal efficiency of TOC was 76.6%. Therefore, although several toxic products were produced, the high TOC efficiency indicated that the concentration of toxic products would be negligible. These results highlight the necessity of tracking toxicity evolution alongside degradation efficiency in wastewater treatment.​