Enhanced Photocatalytic Degradation of Tetracycline over Alcohol-Assisted Bi₂O₃/TiO₂ Composite Heterojunction Under UV Irradiation

Abstract

The widespread presence of antibiotic residues in aquatic environments poses severe ecological risks. While photocatalytic oxidation offers a promising, eco-friendly remediation technology, developing stable and high-efficiency photocatalysts remains a significant challenge. This study investigates the synthesis of Bi₂O₃/TiO₂ heterojunction with tailored morphological structures to enhance the degradation of tetracycline (TC). A series of Bi₂O₃/TiO₂ photocatalysts were prepared via a solvothermal method using mixed alcohol solvents (ethylene glycol and ethanol) to regulate morphology. Comprehensive characterization was performed using XRD, BET, TEM, XPS, UV-Vis, and PL spectroscopy. Photocatalytic activity was evaluated by monitoring TC removal efficiency under light irradiation. The optimized catalyst of BT5-EG3 (n(Bi)/n(Ti) = 0.05; V(EG):V(ethanol) = 1:3) achieved the highest TC conversion of 93.9% within 120 min. This superior performance is attributed to a large specific surface area, abundant lattice oxygen, and a narrowed band gap of 2.52 eV, which significantly promoted the spatial separation of photogenerated charge carriers and suppressed their ultrafast recombination. The reaction followed pseudo-first-order kinetics, and the catalyst demonstrated excellent stability, providing a robust strategy for treating antibiotic-polluted water.

1. Introduction

Antibiotics have been extensively utilized alongside the rapid development of the medical and healthcare sectors, as well as in agriculture, forestry, and animal husbandry [1,2]. However, such antibiotics, veterinary drugs, and pharmaceutical products can readily enter the environment through various pathways, including consumption and improper disposal, leading to severe water pollution. As a result, antibiotic contamination has emerged as a major global environmental concern [3,4,5]. Current treatment strategies mainly fall into three categories: physical, chemical, and biological methods. Physical approaches include adsorption [6], membrane separation [7], and coagulation [8], among others. Biological treatment primarily relies on the metabolic activity of microorganisms such as bacteria, fungi, and algae to break down antibiotics into inorganic compounds [9,10]. Chemical treatment typically involves the generation of highly reactive free radicals with strong oxidative capacity, which non-selectively degrade antibiotics. Photocatalytic advanced oxidation processes represent a promising technique developed on this basis, offering advantages such as high treatment efficiency, low cost, and minimal secondary pollution [11]. Recently, the application of nanomaterials has expanded significantly across various fields, particularly highlighting their paramount importance in environmental remediation and biomedical sectors. Novel eco-friendly approaches, specifically green synthesis methods utilizing plant extracts (e.g., Dracaena reflexa and Elettaria cardamomum), have been rapidly developed to fabricate advanced metal oxide nanoparticles such as TiO₂ and ZnO. These biogenic nanomaterials exhibit exceptional dual-functional capabilities, demonstrating remarkable efficiency in both the photocatalytic degradation of water pollutants and potent antibacterial applications [12,13,14].

Among these, TiO₂ has become the most widely used semiconductor photocatalyst owing to its excellent chemical stability, strong redox capability, corrosion resistance, non-toxicity, and low cost. However, Wu et al. [15] reported that TiO₂ alone removed only 25.1% of tetracycline (TC) under 700 nm light irradiation, and further investigated the corresponding degradation products and mechanism. The practical application of TiO₂ in photocatalysis was nevertheless limited by its narrow spectral response range, low solar light utilization efficiency, and tendency to agglomerate. To enhance its photocatalytic performance, various modification strategies have been adopted, including co-doping with metal ions or non-metal atoms (e.g., sulfur, carbon, and nitrogen) [16,17]. Notably, constructing composite structures or forming heterojunctions with bismuth-based semiconductors can effectively reduce the bandgap energy of TiO₂, promote the separation of photogenerated electron and hole pairs, and significantly improve its oxidation activity [18].

For instance, Hou et al. fabricated a graphene-supported Bi₂O₃ quantum dots/TiO₂ nanosheet composite for degrading rhodamine B dye under visible light in aqueous solution [19]. The composite exhibited markedly higher activity than pure TiO₂ nanosheets, which was attributed to the synergistic effect between TiO₂ and Bi₂O₃ quantum dots. Sood et al. synthesized a heterojunction of anatase TiO₂ and monoclinic α-Bi₂O₃ via a hydrothermal method and evaluated its performance for ofloxacin degradation under natural sunlight [20]. The Bi₂O₃/TiO₂ heterojunction demonstrated superior photocatalytic activity compared to individual TiO₂ or α-Bi₂O₃, achieving complete mineralization of ofloxacin within 120 min, likely due to its well-defined crystalline structure, enhanced visible-light response, and high specific surface area. Wang et al. [21] further improved the photocatalytic degradation of sulfamethazine by developing a two-step calcination method to prepare Bi₂O₃/TiO₂ composites (denoted as Bi-Ti/PAC-2), involving first calcination in air at 300 °C followed by treatment in nitrogen. Under natural sunlight, the degradation efficiency of sulfamethazine in river and lake water using Bi-Ti/PAC-2-700 exceeded 85%, primarily owing to the synergistic action of superoxide radicals, holes, and hydroxyl radicals, with superoxide radicals playing a dominant role. Recently, a novel deep eutectic solvent-assisted route was established for the facile and green synthesis of a (Bi₂O₃/Bi₂O_2.33)/TiO₂ S-scheme heterojunction (BBTO). Compared with pure Bi₂O₃/Bi₂O_2.33 or TiO₂, all BBTO catalysts displayed significantly enhanced activity toward TC degradation, which can be ascribed to the formation of multiple heterojunction interfaces and the presence of oxygen vacancy defects. Moreover, BBTO-0.4 exhibited excellent universality, achieving nearly 100% degradation within 60 min for several other common antibiotics. Based on these findings, a detailed S-scheme charge-transfer mechanism and possible TC degradation pathways were proposed [22].

While the construction of Bi₂O₃/TiO₂ heterojunctions has been proven effective in enhancing the photocatalytic degradation of antibiotics, precisely tailoring their morphological structure and surface oxygen species to further inhibit electron–hole recombination remains a significant challenge. To bridge this gap, this study proposes a facile mixed-solvent (ethylene glycol and ethanol) sol–gel strategy to synthesize highly dispersed Bi₂O₃/TiO₂ heterojunctions with small particle sizes and high specific surface areas. Specifically, during the synthesis, the Bi precursor is converted in situ into highly dispersed Bi₂O₃ nanoclusters that are intimately anchored onto the TiO₂. The coexistence of the Bi₂O₃ and TiO₂ phases, along with the mixed anatase and brookite phases of TiO₂, constructs a highly synergistic Bi₂O₃/TiO₂ heterojunction interface. This tailored approach effectively regulates the microstructure, enriches surface oxygen vacancies, and enhances light absorption.

Therefore, a series of Bi₂O₃/TiO₂ photocatalysts with different n(Bi)/n(Ti) mole ratios and alcohol solvent compositions were systematically prepared. The photocatalytic degradation of tetracycline over these composites under UV irradiation was evaluated, and their structural, optical, and electronic properties were comprehensively characterized using XRD, BET, TEM, UV-Vis, PL, and XPS spectroscopy. The results demonstrated that the optimized BT5-EG3 catalyst exhibited the highest performance owing to its large surface area, abundant oxygen vacancies, narrowed band gap, and significantly suppressed recombination of photogenerated charge carriers. This work provides new insights into the interfacial engineering of Bi₂O₃/TiO₂ heterojunctions for efficient environmental remediation.

2. Experimental

2.1. Preparation of Catalysts

The Bi₂O₃/TiO₂ catalysts with different Bi/Ti molar ratios and alcohol solvents were prepared by sol–gel. The underlying chemistry of this sol–gel process involves the controlled hydrolysis and subsequent polycondensation of the titanium precursor to build a three-dimensional titanium/oxygen framework. First, 13.3 mL of absolute ethanol was added into a beaker and 0.8 g of bismuth nitrate was added into the ethanol accompanied by stirring until completely dissolved. In this step, bismuth nitrate serves as the source for the Bi₂O₃ phases. Then, 11.2 mL of tetrabutyl titanate and 5.2 mL of acetic acid were dropped into the above solution with vigorous stirring for 20 min to obtain a uniform mixed solution. Subsequently, the obtained mixed solutions were aged for 48 h until the gels were formed. This aging period allows for the completion of the polycondensation reactions. After that, the obtained gels were dried at 110 °C for 6 h, and calcined at 450 °C for 4 h, denoted as Bi₂O₃/TiO₂ catalysts (BT). The drying step effectively removes the residual solvents, while the calcination process induces the crystallization of the amorphous gel into the desired photoactive phases. The BT photocatalysts with different molar ratios of Bi/Ti (n(Bi):n(Ti) = 0.01, 0.03, 0.05, 0.07, 0.09) were recognized as BT1, BT3, BT5, BT7, and BT9, respectively. The BT catalysts were synthesized by using different volume ratios of ethylene glycol or glycerol/ethanol (V(ethylene glycol or glycerol)/V(ethanol) = 1:1, 1:3, 1:5, 1:7, 1:9) designated as BT3-EG1, BT3-EG3, BT3-EG5, BT3-EG7, and BT3-EG9, or BT3-G1, BT3-G3, BT3-G5, BT3-G7, and BT3-G9, respectively.

2.2. Performance Evaluation

The performances of BT catalysts for TC degradation were investigated by using a photocatalytic reactor under UV irradiation. A total of 0.5 g of catalysts was dispersed into the 500 mL of 50 mg/L TC solution in a photocatalytic reactor. The TC solution was stirred in darkness for 30 min until adsorption/desorption equilibrium with these photocatalysts was obtained. After that, the photocatalytic degradation of TC was carried out under UV irradiation for 120 min. During the reaction process, the TC solution was taken out, filtered and measured the concentration of TC with the UV-Visible spectrophotometer (355 nm) every 20 min. To accurately determine the exact concentration of TC, a standard calibration curve was established prior to the degradation experiments. A series of TC standard solutions of known concentrations were prepared, and their absorbance at 355 nm was measured. A robust linear relationship between absorbance and concentration (A = kC + b) was obtained, which was then used to translate the measured absorbance into the exact residual concentration (C_t) of the experimental samples at each time interval. The degradation rate was calculated by the following formula [23].

$$E={\displaystyle \frac{C_0-C_{\mathrm{t}}}{C_0}}$$

where E is the TC degradation rate, C₀ is the initial concentration of TC, and C_t is the concentration of TC at different times.

2.3. Characterization

The crystal phases of these photocatalysts were characterized by X-ray diffraction analysis (XRD-6100, Shimadzu Corporation, Kyoto, Japan) with a scan speed of 4°/min using a Cu-target X-ray tube and a diffraction angle (2θ) ranging from 10° to 80°. The specific surface area and pore volume of the catalyst samples were obtained via the BET (ASAP 2460, Micromeritics Instrument Corp., Norcross, GA, USA) and the pore size was computed based on the Barrett–Joyner–Halenda (BJH) method. The microstructures of these samples were examined by transmission electron microscopy (TEM, FEI Company, Hillsboro, OR, USA). The optical properties of these BT catalysts characterized by the UV-Visible spectra (Cary 100, Agilent Technologies, Santa Clara, CA, USA) equipped with an integrating sphere, scanning from 250 to 800 nm with BaSO₄. The excited states of these catalysts’ composite were detected using photoluminescence (MicOS, HORIBA Scientific, Piscataway, NJ, USA). The chemical binding energies of the Bi, Ti, and O were analyzed with the X-ray photoelectron spectrometer (ESCALAB 250XI, Thermo Fisher Scientific, Waltham, MA, USA) using Al Kα radiation at 30.0 eV.

3. Results and Discussion

3.1. Structure of Catalysts

3.1.1. XRD

The XRD patterns of the TiO₂, BT5, and BT5-EG3 catalysts are shown in Figure 1. The diffraction peaks located at 2θ = 25.3°, 37.8°, 48.1°, 53.9°, 55.1°, and 62.7° were assigned to the (101), (004), (200), (105), (211), and (204) crystal planes of anatase TiO₂ (JCPDS No. 84-1286), respectively, and these photocatalysts were mainly identified as anatase TiO₂. However, the characteristic peaks at 25.3°, 25.7°, 30.8°, and 48.1° correspondingly designated to the (210), (111), (211), and (321) planes of brookite TiO₂ (JCPDS No. 01-0737) indicated the strong interaction between Bi₂O₃ and TiO₂. The result suggested that Bi₂O₃ induced the formation of brookite TiO₂. It is noteworthy that no distinct diffraction peaks corresponding to Bi₂O₃ were detected in the XRD patterns. This absence is primarily attributed to the low Bi loading amount (n(Bi)/n(Ti) = 0.05) and the highly dispersed nature of the Bi₂O₃ clusters on the TiO₂ surface, rendering them below the detection limit of conventional XRD. Nevertheless, the presence of Bi species (or Bi₂O₃ phase) was unequivocally confirmed in our subsequent XPS analysis, which corroborates the successful formation of the Bi₂O₃/TiO₂ heterojunction. The mixed phases of TiO₂ (anatase and brookite) were generally considered to show a higher TC degradation activity than pure phase. However, no diffraction peaks of brookite TiO₂ could be detected in the BT5-EG3 sample, with the decrease in intensity of anatase TiO₂ diffraction peaks perhaps attributed to the partial replacement of ethanol by ethylene glycol during the photocatalyst preparation process. From this point of view, it could be deduced that BT5-EG3 exhibited a higher surface area and a lower average particle size owing to the introduction of ethylene glycol.

To quantitatively confirm this deduction and avoid speculative claims, the average crystallite sizes of the synthesized catalysts were calculated using the Scherrer equation according to the following formula based on the prominent anatase (101) diffraction peaks.

$$D={\displaystyle \frac{K \ast \lambda }{\beta \ast \mathit{cos}\left(\theta \right)}}$$

where D is the average crystallite size, K is the shape factor (0.9 for spherical particles), λ is the wavelength of the X-ray used, β is the full width at half maximum (FWHM) of the diffraction peak (in radians), and θ is the Bragg angle (in radians).

The calculated average crystallite size for the optimized BT5-EG3 sample is 14.9 nm, which is significantly smaller than the 23.4 nm calculated for pure TiO₂. This quantitative reduction in crystallite size provides direct structural evidence that the synergistic effect of Bi₂O₃ incorporation and the introduction of mixed alcohol solvents (ethylene glycol) successfully restricts crystal growth. Consequently, this restricted growth directly supports the highly competitive specific surface area (120.0 m²/g) and smaller particle sizes observed in the subsequent BET and TEM analyses.

3.1.2. BET Analysis

The surface area, pore volume, and average pore size of these photocatalysts are summarized in

Table 1. Specific surface areas, pore volumes, and average pore sizes of the photocatalysts.

Samples	SBET (m2/g)	Vp (cm3/g)	Dp (nm)
TiO2	39.2	0.057	5.8
BT1	68.5	0.084	4.9
BT5	79.9	0.084	4.2
BT9	74.3	0.086	4.6
BT5-EG1	121.6	0.133	4.4
BT5-EG3	120.0	0.147	4.9
BT5-EG9	101.9	0.104	4.1
BT5-G3	64.1	0.052	3.2

. The results showed that the BT catalysts all displayed a higher specific surface areas and pore volume than TiO₂, which was beneficial for improving the performance ascribed to the Bi₂O₃ supporting amount. Meanwhile, the specific surface areas of BT samples increased first and then decreased with the increasing Bi₂O₃ supporting amount. Among them, BT5 displayed the highest specific surface area (79.9 m²/g) and the lowest average pore size (4.2 nm), resulting in higher oxidation performance in these BT catalysts. Moreover, compared with BT5, the BT5-EG1, BT5-EG3 and BT5-EG9 samples significantly increased the specific surface areas and pore volume due to the addition of ethylene glycol and the changes in the microstructure. However, for BT5-EG9, the specific surface area decreased with the continuously increasing amount of ethylene glycol. Furthermore, the BT catalysts with the addition of glycerol were not conducive to increasing the specific surface area and improving the oxidation activity. Among these samples, BT5-EG3, with an average pore size of 4.9 nm, showed the highest surface area and pore volume and effectively improved the photocatalytic degradation of TC attributed to the low diffusion resistance and more active sites.

Figure 2a displays the N₂ adsorption/desorption isotherms and pore size distribution of these BT catalysts. All these catalysts, particularly TiO₂, BT5, and BT5-EG3, show typical H4 adsorption/desorption isotherms combined with the H2-hysteresis loop, indicating that these catalysts belong to mesoporous materials, and are characteristic of mesoporous materials’ “inkbottle” pores or the gap among tight spheres. The surface area and pore volume increased remarkably with the Bi₂O₃ supporting amount and the addition of ethylene glycol. Figure 2b shows the pore size distribution curve of these catalysts. It could be found that BT-G3 displayed the lowest average pore size in these samples. However, the BT5-EG3 exhibited a wider range of pore sizes compared to TiO₂ and BT5, which was also beneficial for improving adsorption and oxidation performance.

3.1.3. TEM

The morphology characterizations of the TiO₂, BT5, and BT5-EG3 catalysts were analyzed by TEM. As shown in Figure 3, TiO₂ and BT5 showed an irregular sphere or ellipsoid structure with a particle size of about 20 nm. Furthermore, BT5-EG3 exhibited smaller particle sizes around 12 nm and a higher dispersion than TiO₂ and BT5. The highest surface areas and microstructure of BT5-EG3 were the main reason for the efficient degradation of TC in water by these photocatalysts.

3.2. UV-Vis and PL Spectrum Analysis

To evaluate the optical properties and band gap energy of the synthesized catalysts, the UV-Vis diffuse reflectance spectroscopy of TiO₂ and BT catalysts was studied. As depicted in Figure 4a, compared with TiO₂, BT catalysts with Bi₂O₃ coupling could absorb more visible light due to the red-shifted absorption edges [24,25,26,27]. In particular, BT5-EG1, BT5-EG3 and BT5-EG9 showed that the stronger full-spectrum absorbance than BT attributed to the smaller particle sizes and the higher specific surface area effectively promoted adsorption capacity under UV irradiation. In order to determine the band gap energy (E_g) of these samples, E_g was calculated according to the following formula.

$$E_{\mathrm{g}}=1240/{\lambda }_{\max }$$

As shown in Figure 4b, the band gap energy of TiO₂, BT1, BT5, BT9, BT5-EG1, BT5-EG3, and BT5-EG9 was 3.1, 2.74, 2.52, 2.76, 2.62 eV, 2.52, and 2.52 eV, respectively. The E_g diminished first and then increased with the enhancing Bi₂O₃ supporting amount and BT exhibited the lowest E_g (2.52 eV) in these BT samples. Among them, BT5-EG3 and BT5-EG9, with the highest surface areas following the addition of ethylene glycol, displayed the highest adsorption capacity in the UV-Vis light area, attributed to the lowest band gap energy and structural properties. The outcomes indicated that the enhancement of UV-Vis absorption and the decrease in band gap energy facilitated the movement of photogenerated electron and hole, thereby enhancing the photocatalytic degradation of TC under UV irradiation.

To further highlight the structural and optical advantages of the optimized BT5-EG3 catalyst, a comprehensive comparison with pure TiO₂, as well as other recently reported advanced nanomaterials (such as green-synthesized CuO and ZnO nanoparticles), is summarized in

Table 2. Comparison of crystallite size, particle size, specific surface area, and band gap of the synthesized catalysts with recent reports in the literature.

Catalysts	Synthesis Method	Crystallite Size (nm)	Particle Size (nm)	Surface Area (m2/g)	Band Gap (eV)	Ref.
TiO2	Sol–gel	23.4	~20	39.2	3.1	This work
BT5-EG3	Mixed-solvent sol–gel	14.9	~12	120	2.52	This work
CuO NPs	Green synthesis (E. cardamomum)	11.73	35–600	34.5	2.36	[28]
ZnO NPs	Green synthesis (E. cardamomum)	20.87	9–71	Not reported	3.33	[14]

. Notably, the precisely controlled mixed-solvent sol–gel strategy employed in this study yields BT5-EG3 with a uniquely small crystallite size and a highly competitive specific surface area (120.0 m²/g). This surface area is significantly larger than that of pure TiO₂ (39.2 m²/g) and recently reported phyto-engineered CuO nanomaterials (34.5 m²/g) [28]. Furthermore, the narrowed band gap of BT5-EG3 (2.52 eV) demonstrates excellent visible-light responsiveness compared to pure TiO₂ (3.10 eV) and biogenic ZnO (3.33 eV) [14]. These comparisons clearly confirm that the synergistic effect of the heterojunction of Bi₂O₃ and TiO₂ and mixed-solvent regulation successfully tailors the morphological structure and electronic properties of the catalyst, thereby providing abundant active sites for enhanced photocatalytic performance.

The PL spectrum is often utilized to explore the charge transfer efficiency of photocatalysts, which can reveal the separation and recombination rate of photogenerated electron and hole. As depicted in Figure 5, the PL spectra exhibited the emission bands located at 650–800 nm for TiO₂, BT5 and BT5-EG3, suggesting that Bi₂O₃ introduction or the addition of ethylene glycol did not produce extra emission signal and only changed the intensity. Compared with TiO₂ and BT5, BT5-EG3 exhibited a significant quenching of the PL emission intensity, which unambiguously indicates a substantially suppressed radiative recombination of photogenerated electron–hole pairs. This is primarily attributed to the efficient spatial charge separation across the intimate contact interface of the Bi₂O₃/TiO₂ heterojunction. Furthermore, the addition of ethylene glycol optimized the microstructure, thereby prolonging the lifetime of the charge carriers and enhancing the ultimate photodegradation activity for TC.

3.3. XPS Analysis

In order to further investigate the effects of ethylene glycol or glycerol replacement on the element composition and electron state of Bi₂O₃/TiO₂, XPS was performed to characterize the binding energy of the Ti 2p, Bi 4f, O 1s of the BT5-EG3 and BT5-G3 samples (Figure 6). The peaks located at about 458.8 and 464.6 eV corresponded to Ti 2p_1/2 and Ti 2p_3/2 spin-orbit splitting of Ti⁴⁺, separately [29]. To further elucidate the strong interfacial electronic interactions within the heterojunction, a high-resolution peak deconvolution of the Ti 2p spectra was performed. The results showed that the signal intensity of Ti⁴⁺ of BT5-EG3 was significantly higher than BT5-G3, indicating that more Ti⁴⁺ active sites of BT5-EG3 were exposed than BT5-G3. Consequently, the overall binding energy of Ti 2p of BT5-EG3 shifts toward a lower energy direction compared to BT5-G3, confirming the increased electron cloud density around Ti atoms induced by this robust interfacial electronic coupling of Bi and Ti [30]. Figure 6b shows the XPS spectrum of Bi 4f, and the peaks at around 159.4 and 164.7 eV could be attributed to Bi 4f_7/2 and Bi 4f_5/2, respectively. Similarly, BT5-EG3 was shifted in the direction of the lower binding energy relative to BT5-G3 [31], indicating a stronger electronic interaction at the Bi₂O₃/TiO₂ interface in BT5-EG3. The O 1s spectra were not symmetrical for both the BT5-EG3 and BT5-G3 samples, implying the coexistence of several oxidation states of oxygen atoms on the surface. The fitted peaks of O 1s were deconvoluted into three components: lattice oxygen (O_Latt), oxygen vacancies (O_V), and surface adsorbed oxygen/hydroxyl groups (O_Ads) [32]. To quantitatively analyze the variations in surface oxygen species induced by the heterojunction formation and solvent regulation, the relative ratios of these oxygen species were calculated based on their integrated peak areas (Figure 6c). The quantitative analysis demonstrates a significant evolution in the oxygen composition. For the BT5-G3 sample, the proportions of O_Latt, O_V, and O_Ads are 60.9%, 26.3%, and 12.9%, respectively. In contrast, the optimized BT5-EG3 sample exhibits a substantial decrease in lattice oxygen (42.2%) and a concomitant significant increase in both oxygen vacancies (29.4%) and adsorbed oxygen species (28.4%). The generation of these abundant oxygen vacancies is an inherent interfacial defect response driven by the intimate contact within the Bi₂O₃ and TiO₂ heterojunction, as well as the localized reducing environment provided by ethylene glycol during the calcination process. Therefore, BT5-EG3 possesses a much higher surface oxygen adsorption capacity and oxygen vacancy concentration than BT5-G3. This significantly facilitates the robust generation of reactive radical species (·O₂⁻, ·OH), thereby boosting the photocatalytic degradation efficiency of TC. It can be concluded that BT5-EG3 exhibits optimized surface structural characteristics, enhanced interfacial coupling, and more efficient separation of photogenerated electron–hole pairs [33]. Furthermore, this high concentration of reactive surface oxygen species (O⁻, ·OH) and defect sites significantly facilitates the robust generation of radical species, thereby boosting the photocatalytic degradation efficiency of TC. Therefore, BT5-EG3 possessed a higher surface oxygen adsorption amount (O⁻, ·OH) and higher oxygen vacancy concentration than BT5-G3 with a slight shift toward lower binding energy. Meanwhile, the high concentration of surface oxygen species (O⁻, ·OH) favored the oxidation activity of TC degradation.

3.4. Photocatalytic Performance

The photocatalytic degradation of TC over various BT catalysts is shown in Figure 7. It showed the photocatalytic degradation of TC over TiO₂, BT1, BT2, BT3, BT4, BT5, BT6, and BT9 catalysts at different reaction times. It could be seen that the adsorption/desorption equilibrium for all samples was obtained in darkness after 30 min, and then the concentration of TC remained relatively stable. It could be also found that the oxidation activities of BT series catalysts were all higher than TiO₂ after illumination for 120 min. As the Bi₂O₃ supporting amount of these BT catalysts continuously increased, the degradation efficiency for TC first increased and then decreased. Among them, BT5 showed the highest degradation rate (92.3%), resulting in a final TC concentration of 2.6 mg/L, while it was only 58.4% for TiO₂. The higher oxidation performance of BT catalysts benefited from the higher specific surface area. Meanwhile, the strong interfacial coupling between Bi₂O₃ and TiO₂ produced new modification energy levels, enhanced UV-Vis light absorbance, and reduced band gap energy with the Bi₂O₃ supporting amount. In addition, the robust formation of the Bi₂O₃/TiO₂ heterojunction induced a strong synergistic effect that significantly facilitated the interfacial migration and spatial separation of photogenerated electron–hole pairs, thereby effectively retarding their recombination and prolonging the lifetime of the reactive charge carriers. This was consistent with the results obtained from XRD, BET, and UV-Vis characterization.

In order to further promote the oxidation activity of BT5 catalysts, BT5-EG and BT5-G prepared with the addition of ethylene glycol or glycerol were also investigated for photocatalytic degradation of TC (Figure 8). As shown in Figure 8a, the TC degradation performance increased first and then decreased subsequently for BT5-EG photocatalysts with an increasing amount of ethylene glycol, even worse than BT5, including the BT5-EG7 and BT5-EG9 samples. Among them, BT5-EG3 indicated the highest oxidation performance (93.9%) owing to the high specific surface area, amount of adsorbed oxygen and lattice oxygen, and modification of the microstructure with the addition of ethylene glycol. In contrast, as displayed in Figure 8b, all BT5-G samples showed lower oxidation efficiency than BT5, maybe due to the structure and shape change resulting in the low surface area, pore volume/size, and reduced content of active sites on the surface. In conclusion, the difference of structural characteristics of BT5-EG and BT5-G with the addition of ethylene glycol or glycerol was considered to be mainly responsible for the variation in adsorption properties and photocatalytic performances of TC removal.

Simultaneously, photocatalytic degradation and pseudo-first-order kinetics modeling of various initial concentrations of TC over BT5-EG3 is shown in Figure 9. It could be found that the degradation efficiency and speed decreased gradually with the increasing initial concentrations of TC from 10 to 50 mg/L due to the high TC concentration affecting the adsorption of these catalysts and occupying the active sites during the adsorption process. Furthermore, the catalysts of the saturated state greatly influenced the degradation rate of TC. Therefore, the photocatalytic active sites were limited and blocked under the high concentration of TC. Therefore, the high initial concentrations of TC required a greater amount of catalysts or longer time to achieve complete degradation.

To emphasize the uniqueness and superior performance of the proposed method, the experimental parameters and degradation results of the synthesized Bi₂O₃/TiO₂ heterojunction were compared with recent studies involving TiO₂, ZnO, and other metal oxide-based photocatalytic systems (

Table 3. Comparison of photocatalytic degradation parameters of the present catalyst with recently reported photocatalysts.

Photocatalyst	Target Pollutant (Conc.)	Catalyst Dose	Light Source (Wavelength)	Time (min)	Removal Efficiency (%)	Kinetic Rate	Ref.
Bi2O3/TiO2	Tetracycline (50 mg/L)	1.0 g/L	UV light	120	93.9	Pseudo-first-order	This work
CdS/porous ZnO-PEG	Methylene blue (5 mg/L)	1 disk (0.5 cm2)	UV lamp (250 W)	30	80.85	-	[34]
Zn-doped BiOBr	Tetracycline (10 mg/L)	0.2 g/L	Visible LED (60 W)	180	75.94	-	[35]
TiO2@S-AZMB/ZnTi-LDH	Metronidazole (20 mg/L)	1.5 g/L	Xenon lamp (300 W)	180	94	0.0191 min−1	[36]
SiO2@TiO2 (with PMS)	ortho-Phenylphenol (17 mg/L)	1.0 g/L	UV-A (352 nm)	120	95	-	[37]
Ce-Ag-ZnO	Reactive red 120 (20 mg/L)	0.1 g/L	UV lamp	60	>99.0	Pseudo-first-order	[38]
Fe-doped BiOBr	Crystal violet (20 mg/L)	-	Visible light	60	96.8	-	[39]
TiO2-x/FeTiO3	Tetracycline (~27.5 mg/L)	0.5 g/L	Visible LED (50 W)	-	High	-	[40]
ZnO/N-Cu-MOF	H. perezi (106 cells/mL)	1.01 mg/L	Visible (400–700 nm)	120	IC50 (Inactivation)	-	[41]
Cu/TiO2	Milled cellulose	-	UV light	-	H2 evolution	-	[42]
Bi2O3 NPs	Atrazine (5 mg/L)	0.8 g/L	UV-A light	60	92.1	-	[43]

). Compared to other photocatalysts that often require either higher catalyst dosages, extended irradiation times, or additional chemical promoters, the alcohol-assisted Bi₂O₃/TiO₂ composite developed in this work demonstrates a highly competitive removal efficiency (93.9%) for a relatively high concentration of tetracycline (50 mg/L) within 120 min. This substantial activity can be attributed to the synergistic effect of the heterojunction formed between Bi₂O₃ and TiO₂, which significantly accelerates the pseudo-first-order kinetic rate and promotes effective charge carrier separation under UV irradiation.

Figure 10 shows the photocatalytic degradation stability of TC over BT5-EG3 with three recycles. It was easy to find that the degradation rate over BT5-EG3 declined slowly, mainly attributed to the reduced adsorption in repeated reaction. In addition, the TC adsorbed onto the surface of the catalyst reduced the light-harvesting ability, resulting in the increasing of electron–hole recombination. Therefore, it required a longer reaction time for BT5-EG3 to reach above 90% degradation efficiency.

3.5. Photocatalytic Oxidation Mechanism

To identify the main active species, isopropanol (IPA, 5 mM), benzoquinone (BQ, 0.2 mM), and ammonium oxalate (AO, 5 mM) were added to the photocatalytic reaction as the scavengers of ·OH, ·O₂⁻, and h⁺, respectively. The effect of different scavengers for photocatalytic degradation of TC over BT5-EG3 is shown in Figure 11. The results demonstrated that the degradation efficiency of TC over BT5-EG3 was 92.3% without the scavenger. However, the degradation efficiencies decreased by 9.3%, 30.2%, and 15.6% with the adding of IPA, BQ, and AO, respectively, indicating that ·O₂⁻ was the dominant active species for TC degradation over BT5-EG3. Furthermore, it revealed a significant decrease in the TC photodegradation rate under nitrogen protection without oxygen. The results further confirmed that ·O₂⁻ played an important role during the TC degradation process.

Based on related reports [15,21] and the above experimental results, the possible reactions involved are as follows:

$$\begin{array}{l}BT+\mathrm{h}\nu \to {\mathrm{h}}^{+}(Ti{O}_2{)}_{(VB)}+e(Bi{)}_{(CB)}\\ h^{+}(Ti{O}_2{)}_{(VB)}+H_{2}O\to •OH+H^{+}\\ e(Bi{)}_{(CB)}+O_2\to •O_2^{-}\\ •O_2^{-}+H^{+}\to H{O}_2•\\ 2H{O}_2•\to H_2{O}_2+O_2\\ TC+•O_2^{-}\to C{O}_2+H_{2}O\end{array}$$

The possible mechanism of photocatalytic degradation of TC over the BT5-EG3 composite is presented in Figure 12. The enhanced photocatalytic performance of the BT5-EG3 composite is primarily attributed to the construction of a robust heterojunction architecture between the Bi₂O₃ and TiO₂ phases, which facilitates the spatial separation of photogenerated electron–hole pairs. During the synthesis, the Bi precursor crystallizes into highly dispersed Bi₂O₃ phases that form an intimate contact with the TiO₂. This large contact interface between the two phases is highly beneficial for accelerating charge transfer and separation. Furthermore, this phase coexistence promotes a reduction in the band gap energy to 2.52 eV. Under UV light irradiation, both Bi₂O₃ and TiO₂ are excited to generate electron–hole pairs. Based on the staggered band alignment characteristic of a conventional type II heterojunction, the conduction band (CB) and valence band (VB) levels of Bi₂O₃ are higher than the corresponding levels of TiO₂. Driven by this specific band alignment, the photogenerated electrons on the CB of Bi₂O₃ thermodynamically migrate to the CB of TiO₂, while the photogenerated holes migrate to or remain on the VB of Bi₂O₃. This spatial separation process effectively accelerates the migration of charge carriers across the heterojunction interface and overcomes the ultrafast electron–hole recombination typically observed in single semiconductors. Consequently, the spatially separated electrons accumulated on the CB of TiO₂ are captured by adsorbed O₂ to produce superoxide radicals (·O₂⁻), and H₂O molecules accept holes (h⁺) to generate ·OH radicals. Ultimately, the TC molecules adsorbed onto the surface of BT5-EG3 are successively oxidized under the synergistic action of these highly reactive radical species (·O₂⁻ and ·OH).

While the BT5-EG3 catalyst demonstrated excellent efficiency in degrading TC under UV irradiation, a limitation of this study is the reliance on a single antibiotic model in a controlled matrix. The HRTEM image for resolved fringes and SAED pattern will be further researched to master the structural morphology of BT5-EG3 catalysts. Simultaneously, the catalytic oxidation mechanism will be investigated using LC-MS, in situ IR, Mott–Schottky and VB-XPS analyses, electrochemical impedance spectroscopy (EIS) and so on. Additionally, real-world aquatic environments contain complex matrices and co-existing pollutants that may interfere with the photocatalytic process. Future work will focus on evaluating the performance of these heterojunctions under natural sunlight and in actual wastewater matrices.

4. Conclusions

In this work, a series of Bi₂O₃/TiO₂ composites catalysts prepared by the sol–gel method with different Bi/Ti molar ratio and alcohol solvent have been studied for photocatalytic degradation of TC under UV irradiation. Among these photocatalysts, BT5-EG3 displayed the highest oxidation activity (93.9%) for TC removal due to the higher surface area, adsorption (BET), and amount of adsorbed and lattice oxygen (XPS) attributed to the different structure with the assistance of alcohol solvents. According to the results of the analysis, the introduction of Bi species promoted the in situ growth of highly dispersed Bi₂O₃ nanoclusters on the TiO₂ surface, resulting in the formation of robust Bi₂O₃/TiO₂ heterojunctions and the partial induction of brookite phase. This unique structure, coupled with the lower bandgap energy (2.52 eV) was highly beneficial for optimizing the band alignment and facilitating the spatial separation of electron–hole pairs. Consequently, as corroborated by the PL spectra, the recombination of charge carriers in the BT5-EG3 composite was remarkably retarded, thereby maximizing the generation and utilization of highly reactive radical species (·OH and ·O₂⁻) for efficient TC degradation. The potential photocatalytic oxidation mechanism demonstrated that ·O₂⁻ was the main active species during the photocatalytic reaction process. The oxidation ability, stability and pseudo-first-order kinetics model of BT5-EG3 were also investigated under the different initial concentrations of TC. The present investigation provided a useful strategy for the design of efficient TiO₂ photocatalysts and new insights for controlling antibiotic pollution in water.