Sustainable Photocatalytic Treatment of Real Pharmaceutical Wastewater Using a Novel ZnO/MIP-202(Zr) Bio-MOF Hybrid Synthesized via a Green Approach

Abstract

Metal–organic frameworks (MOFs) are promising materials for environmental remediation, particularly in photocatalysis. In this work, a novel ZMIP nanocomposite was fabricated by integrating MIP-202(Zr) bio-MOF with ZnO nanoparticles. For the first time, ZnO nanoparticles were green-synthesized using water lettuce extract and incorporated into MIP-202(Zr) via a mild hydrothermal route. The resulting hybrid was applied as a visible-light photocatalyst for carbamazepine (CBZ) degradation in real pharmaceutical wastewater. Structural analyses (XRD, FTIR, TEM, EDS) verified the successful incorporation of ZnO into the MIP-202(Zr) framework. The composite exhibited a narrowed bandgap of 2.74 ± 0.1 eV compared to 4.05 ± 0.06 eV for pristine MIP-202 and 3.77 ± 0.04 eV for ZnO, highlighting enhanced visible-light utilization in ZMIP. Operational parameters were optimized using response surface methodology, where CBZ removal reached 99.37% with 84.39% TOC mineralization under the optimal conditions (90 min, pH 6, 15 mg/L CBZ, 1.25 g/L catalyst). The catalyst maintained stable performance over five reuse cycles. Radical quenching and UHPLC-MS analyses identified the dominant reactive oxygen species and generated intermediates, elucidating the degradation mechanism and pathways. Beyond CBZ, the ZMIP photocatalyst effectively degraded other pharmaceuticals, including doxorubicin, tetracycline, paracetamol, and ibuprofen, achieving degradation efficiencies of 82.93%, 76.84%, 72.08%, and 67.71%, respectively. Application on real pharmaceutical wastewater achieved 78.37% TOC removal under the optimum conditions. Furthermore, the supplementation of the photocatalytic system by inorganic oxidants ameliorated the degradation performance, following the order KIO₄ > K₂S₂O₈ > KHSO₅ > H₂O₂. Overall, ZMIP demonstrates excellent activity, reusability, and versatility, underscoring its potential as a sustainable photocatalyst for real wastewater treatment.

1. Introduction

The accelerated production and consumption of drugs for overcoming diseases in humans and animals have resulted in the detection of refractory pollutants in water streams [1]. Carbamazepine (CBZ), an antiepileptic and psychotropic drug, is commonly found in water systems due to its massive global consumption (1000 tons/year) [2]. Further, CBZ can reach water systems in different concentrations from ng/L to µg/L through wastewater treatment plants, livestock farming, and hospitals’ wastewater [3]. Hai et al. [4] mentioned that the CBZ concentration ranged from 30 to 6300 ng/L in the effluent of wastewater treatment plants in different countries such as Canada, Australia, and Finland. The accumulation of CBZ in water sources due to its bio-recalcitrance can harm aquatic organisms such as fish, algae, and invertebrates [5]. To mitigate the release of CBZ into the environment, various conventional biological, chemical, and physical treatment methods have been employed, including bacterial and fungal biodegradation, adsorption, electrocoagulation, and membrane filtration. However, their limited efficiency, high energy demand, high cost, and the generation of secondary pollution hinder their practical application [6]. Therefore, the development of an efficient, simple, cost-effective, and environmentally friendly degradation system is essential to mitigate the release of CBZ into aquatic environments.

Advanced oxidation processes (AOPs) such as Fenton, ozonation, and photocatalysis have been recently utilized for the degradation of bio-resistant pollutants (e.g., drugs) [7,8,9]. However, Fenton process necessitates acidic conditions and produces sludge, while ozonation process is expensive and generates toxic by-products. Regarding photocatalysis, it is a promising technique due to its simplicity, effectiveness, inexpensiveness, and green nature [10]. In photocatalysis, electron-hole (e⁻/h⁺) pairs can be generated through the excitation of a semiconductor (e.g., TiO₂, ZnO) by a light source with suitable energy leading to the production of reactive oxygen species (ROS) such as hydroxyl radicals (^•OH) and superoxide radicals (O₂^{• −}) and that can effectively degrade organic pollutants [11]. ZnO is featured by its high chemical stability, environmental friendliness, availability, low-cost, large excitation binding energy, and high electron mobility; therefore, it is a promising metal oxide for the employment in different applications such as CO₂ reduction, water splitting, and photocatalytic degradation [12]. However, the preparation of ZnO requires the consumption of toxic chemicals which harms the environment and obstructs the scalable production. Thus, in this study, plant extracts were used as a reducing and stabilizing agents to replace dangerous chemicals which facilitates the practical production of ZnO. Water lettuce plant was used to fabricate ZnO due to its availability in water streams. The application of water lettuce in the green synthesis of ZnO can contribute to solving different environmental issues related to water lettuce such as blocking water streams, depletion of oxygen, increasing floods, and inhibiting biodiversity [13]. On the other hand, the large optical band gap (E_g) of ZnO and its inefficient charge carrier separation significantly limit its scalable application for the degradation of organic micropollutants, thereby hindering its photocatalytic performance [14]. These defects can be managed by different approaches such as metal doping, non-metal doping, fabricating heterojunctions, and decorating noble metals [15]. The fabrication of heterojunction can ameliorate the photocatalytic performance due to the charge transfer between the constituents of the heterostructure leading to the improvement of charge carrier separation. Oliveira et al. [16] synthesized a Bi₂MoO₆/ZnO heterojunction for the degradation of methylene blue dye indicating the improvement of the degradation efficiency to 99.7% in the case of the heterojunction compared to around 80% for pure ZnO. Nonetheless, the construction of heterojunction between ZnO and metal–organic frameworks (MOFs) has been rarely explored in the literature, and the degradation mechanism remains poorly investigated.

Recently, MOFs which are made of metal ions and organic ligands have been employed as photocatalysts due to their unique properties such as high surface area, porosity, and stability [17,18]. However, MOFs have some defects such as wide E_g, fast reunite of electron-hole pairs, and high cost due to the need for harmful, high-cost chemicals and high temperatures during the synthesis [19]. In this study, a zirconium-based bio-metal–organic framework (MIP-202(Zr) bio-MOF) was synthesized, where L-aspartic acid was employed as an organic ligand without using harmful chemicals and under low temperature [20]. However, the photocatalytic performance of MIP-202 is restricted due to its wide E_g and fast recombination rate. The integration of ZnO with MIP-202 can effectively suppress charge carrier recombination through interfacial transfer of electron–hole pairs between the two materials. Moreover, this interaction may lead to the formation of new energy levels during composite fabrication, thereby extending light absorption into the visible region. These synergistic effects enhance degradation kinetics and contribute to achieving higher mineralization efficiency [21]. Zhang et al. [22] fabricated a heterojunction of ZnO@NH₂-MIL-88B showing the decrease in E_g to 2.22 eV in the case of ZnO@NH₂-MIL-88B compared to 3.19 eV in the case of pure ZnO and the improvement of charge carrier separation which resulted in the enhancement of degradation performance for ZnO@NH₂-MIL-88B compared to pure ZnO. In a recent study, C- and N-co-doped ZnO photocatalysts were synthesized via pyrolysis using ZIF-8 as the precursor. Optical characterization revealed that the band gap of pristine ZnO was approximately 3.37 eV, whereas the co-doped ZnO (C/N-ZnO-500) exhibited a reduced band gap, indicative of enhanced absorption within the visible-light region [23]. The coupling between ZnO and MIP-202 for the application in photocatalytic degradation of organic pollutants has not been explored in the previous studies, and the investigation of the degradation mechanism can provide new insights into the literature.

In this study, a novel ZnO/MIP-202(Zr) (ZMIP) heterostructure was fabricated by integrating bio-derived MIP-202(Zr) with ZnO nanoparticles through a mild hydrothermal process. To the best of our knowledge, this represents the first report on the construction of such a heterostructure. Furthermore, ZnO nanoparticles were synthesized via a green and sustainable route using water lettuce extract, providing an eco-friendly alternative to conventional chemical synthesis. This dual innovation highlights both the novelty in material design and the cost-effective, sustainable direction of the proposed photocatalyst. The synthesized ZMIP photocatalyst was thoroughly characterized using XRD, FTIR, TEM, and EDS to validate its structural integrity, morphological features, and elemental composition, thereby confirming the successful incorporation of ZnO into the MOF. In addition, the optical properties, including the band gap (Eg) and the relative positions of the conduction band (CB) and valence band (VB), were examined to assess the material’s light absorption capacity and charge transfer potential. The photocatalytic activity of ZMIP was then systematically investigated under visible light irradiation using CBZ as a model pharmaceutical contaminant. The influence of operational parameters such as irradiation time, pH, initial CBZ concentration, and catalyst dosage was optimized through response surface methodology (RSM), ensuring a statistically robust evaluation of the system. The kinetics of CBZ degradation were examined by applying pseudo-first-order and pseudo-second-order models, while total organic carbon (TOC) mineralization was assessed to evaluate the extent of organic load reduction as an additional indicator of the treatment performance. Catalyst durability and reusability were further studied through multiple repeated photocatalytic cycles. The underlying degradation mechanism was explored by combining optical property analysis with radical quenching experiments to clarify charge transfer pathways and identify the dominant reactive oxygen species, respectively. In addition, the degradation pathways were elucidated by detecting and characterizing transformation intermediates using an ultra-high-performance liquid chromatography coupled with mass spectrometry (UHPLC-MS) to provide insights into the sequential breakdown of CBZ. To explore its versatility and practical potential, the ZMIP/light system was further applied to the degradation of structurally diverse pharmaceuticals (doxorubicin hydrochloride, tetracycline, paracetamol, and ibuprofen) and tested across different water matrices (deionized water, tap water, lake water, seawater, and drain water) and real industrial wastewater collected from untreated pharmaceutical discharge. In addition, the influence of typical coexisting constituents, including natural organic matter (NOM, e.g., humic acid), inorganic anions (Cl⁻, HCO₃⁻, NO₃⁻, SO₄²⁻), and inorganic cations (Mg²⁺, Ca²⁺, Na⁺, Cu²⁺), on CBZ degradation within the ZMIP/light system was systematically examined. Finally, the incorporation of highly reactive inorganic oxidants was examined to evaluate their effectiveness in mitigating electron–hole recombination and promoting the generation of reactive oxygen species. The tested inorganic oxidants were hydrogen peroxide (H₂O₂), potassium monopersulfate (KHSO₅), potassium peroxydisulfate (K₂S₂O₈), and potassium periodate (KIO₄). Collectively, this work provides the first demonstration of a green-synthesized ZnO/MIP-202(Zr) heterostructure as a multifunctional, durable, and sustainable photocatalyst, supported by comprehensive experimental investigations ranging from synthesis and characterization to mechanistic studies and real-world applicability.

2. Results and Discussion

2.1. Physicochemical Characteristics of the Synthesized Materials

2.1.1. X-Ray Diffraction (XRD)

Figure 1a presents the XRD pattern of the green-synthesized ZnO nanoparticles, which shows excellent agreement with the reference data for the hexagonal wurtzite phase of zinc oxide (JCPDS No. 00-36-1451) [24]. Distinct diffraction signals were observed at 2θ angles of 31.69°, 34.36°, 36.17°, 47.53°, 56.52°, 62.80°, 66.30°, 67.97°, and 69.38°, corresponding to the (100), (002), (101), (102), (110), (103), (200), (112), and (201) crystallographic planes, respectively. These reflections confirm the successful synthesis of ZnO with a well-defined crystalline wurtzite structure. As shown in Figure 1b, the XRD pattern of the synthesized MIP-202(Zr) bio-MOF exhibits distinct diffraction peaks at 2θ values of 8.63°, 9.74°, 19.73°, and 21.59°, which can be indexed to the (111), (200), (420), and (440) crystallographic planes, respectively [25]. The sharp and well-defined reflections confirm the successful synthesis of the MIP-202(Zr) framework with a high degree of crystallinity. Additionally, the obtained XRD pattern of the synthesized MOF was compared with literature-reported data, as summarized in Table S1. The strong agreement between the observed and reported diffraction peaks verifies that the synthesized material possesses the characteristic crystal structure of MIP-202(Zr). On the other hand, the XRD pattern of the ZMIP photocatalyst (Figure 1c) displays characteristic diffraction peaks at 2θ values of 31.88°, 34.64°, 36.04°, 47.92°, 56.23°, 62.52°, 66.40°, 68.80°, and 69.70°, which are assigned to the (100), (002), (101), (102), (110), (103), (200), (112), and (201) planes of ZnO. Additional reflections at 8.12°, 9.74°, 13.27°, 20.14°, and 21.48° are attributed to the (111), (200), (222), (420), and (440) planes of MIP-202(Zr). The coexistence of diffraction features from both ZnO and MIP-202(Zr) demonstrates the successful formation of the ZMIP heterostructure, confirming the effective incorporation of the bio-MOF with the ZnO nanoparticles. Furthermore, for clarity and ease of comparison, all XRD patterns have also been compiled into a single combined figure (Figure S1).

2.1.2. Fourier-Transform Infrared (FTIR) Spectroscopy

Figure 2 presents the FTIR spectrum of the ZnO nanoparticles, MIP-202(Zr) bio-MOF, and ZMIP nanocomposite. In the FTIR spectrum of the synthesized ZnO nanoparticles, a characteristic absorption peak at 471.98 cm⁻¹ is assigned to the Zn–O stretching vibration [26]. This peak confirms the successful formation of ZnO nanoparticles. In addition, the spectrum displays a broad and intense band at 3412.88 cm⁻¹, which is attributed to the O–H stretching vibrations [27]. Further absorption features are observed at 2920.96 and 1408.69 cm⁻¹, corresponding to C–H stretching and bending vibrations, respectively [28]. The band located at 1584.83 cm⁻¹ is ascribed to aromatic C–C stretching [29], while the peak at 1049.58 cm⁻¹ may be related either to C–N stretching of primary amines or to C–O stretching of primary alcohols [29].

The FTIR spectrum of the synthesized MIP-202(Zr) bio-MOF exhibits a broad absorption band centered at 3204.62 cm⁻¹, which is attributed to the asymmetric and symmetric stretching vibrations of the −NH₂ groups originating from L-aspartic acid [30]. The bands observed at 1610.3 and 1423.6 cm⁻¹ correspond to the asymmetric and symmetric stretching modes of C−O bonds within the bio-organic linker, respectively [31]. A distinct absorption at 1234.77 cm⁻¹ is attributed to C−N stretching vibrations [25]. Furthermore, the peaks at 647.61 and 482.27 cm⁻¹ are assigned to Zr−COO and Zr−O bonds, respectively [31]. Collectively, these characteristic vibrational features provide compelling evidence for the successful assembly of the MIP-202(Zr) bio-MOF.

The FTIR spectrum of the ZMIP nanocomposite displays a broad absorption band at 3015.94 cm⁻¹, which is attributed to the O–H stretching vibrations of the green-synthesized ZnO nanoparticles. This band partially overlaps with the −NH₂ stretching modes arising from the MIP-202(Zr) bio-MOF, indicating successful integration of the two components within the composite structure. The band at 1600.47 cm⁻¹ corresponds to the asymmetric C–O stretching vibration of the bio-organic linker, while the absorption at 1505 cm⁻¹ is assigned to aromatic C–C stretching in ZnO. In addition, the signal observed at 1419 cm⁻¹ is associated with C–H bending vibrations of ZnO, which overlaps with the symmetric C–O stretching of the MIP-202(Zr) linker. A distinct band at 1250.17 cm⁻¹ is ascribed to C–N stretching vibrations within the MIP-202(Zr) framework, and the peak at 988.63 cm⁻¹, slightly shifted, may correspond to either C–N stretching of primary amines or C–O stretching of primary alcohols in ZnO. Moreover, the absorption at 653.72 cm⁻¹ is consistent with Zr–COO bonding in the MIP-202(Zr) MOF, whereas the band at 463.94 cm⁻¹ arises from the combined contributions of Zn–O vibrations in ZnO and Zr–O bonds in MIP-202. The slight shifts in peak positions relative to the individual components are attributed to the incorporation of ZnO nanoparticles into the MOF, which modifies the local electronic environment and influences hydrogen-bonding interactions of amine groups [32]. The FTIR results confirm the coexistence of characteristic functional groups from both components, validating the successful synthesis and effective integration of the ZMIP heterostructure.

2.1.3. Structural Morphology and Elemental Profiling

The transmission electron microscopy (TEM) micrographs of the green-synthesized ZnO nanoparticles at different magnifications (Figure 3a,b) revealed a predominantly spherical morphology with uniform spatial distribution. The particles exhibited an average diameter of approximately 9.06 nm, as further confirmed by the particle size distribution shown in Figure S2. The high-resolution TEM (HRTEM) image (Figure 3c) further confirmed their crystalline nature, as evidenced by distinct lattice fringes corresponding to the interplanar spacings characteristic of the hexagonal wurtzite phase of ZnO. This observation verifies the successful synthesis of ZnO nanoparticles. Moreover, the energy-dispersive X-ray spectroscopy (EDS) spectrum (Figure 3d) indicated a surface composition of 81.61 wt% Zn and 18.39 wt% O, consistent with the theoretical stoichiometry of ZnO. Complementary elemental mapping images (Figure 3e) demonstrated a homogeneous and well-integrated distribution of Zn and O throughout the nanoparticle matrix, confirming both compositional uniformity and structural integrity.

The TEM images of the MIP-202(Zr) bio-MOF at various magnifications (Figure 4a,b) revealed that the particles possessed a primarily spherical geometry with nanoscale features. The average particle diameter was estimated to be approximately 70.94 nm, as further confirmed by the particle size distribution presented in Figure S3. While the XRD profile of MIP-202 (Figure 1b) confirms its overall crystallinity, the HRTEM image (Figure 4c) reveals the presence of locally disordered or amorphous regions. Such features are frequently observed in MOFs and are often attributed to partial amorphization caused by electron beam irradiation, surface or defect-related heterogeneities, and the intrinsic dynamic transition between crystalline and amorphous domains [33]. Moreover, the EDS spectrum (Figure 4d) and elemental mapping (Figure 4e) of the MIP-202(Zr) bio-MOF provide further confirmation of its successful synthesis and uniform elemental composition. As shown in the EDS spectrum, the framework is mainly composed of Zr (43.8 wt%), C (34.18 wt%), O (11.5 wt%), Cl (8.86 wt%), and N (1.65 wt%). The elevated levels of Zr, C, and O are consistent with the presence of Zr-cluster nodes and bio-organic linkers, which constitute the fundamental structural units of the MIP-200 MOF. The detected Cl is attributed to residual chloride ions retained within the porous framework [31], whereas the nitrogen content originates from the −NH₂ groups of the aspartic acid linker. Moreover, the elemental mapping images of the bio-MOF illustrate the homogeneous spatial distribution of Zr, C, O, Cl, and N, with the merged map, confirming their uniform dispersion across the framework. These observations indicate that the MIP-202(Zr) bio-MOF possesses a well-organized structure in which the key elements are evenly integrated, ensuring structural integrity and functional performance.

The TEM micrographs of the ZMIP nanocomposite reveal particles with irregular morphology and a distinctly porous architecture, as shown in Figure 5a,b. The HRTEM image (Figure 5c) further highlights the pronounced interfacial interactions between the ZnO nanoparticles and the MIP-202(Zr) bio-MOF, thereby confirming their effective integration at the nanoscale. The EDS spectrum (Figure 5d) indicates that the nanocomposite consists of Zr (62.09 wt%), C (11.29 wt%), Cl (7.7 wt%), and N (2.16 wt%) derived from the MIP-202(Zr) framework, while the detection of Zn (2.69 wt%) validates the incorporation of the ZnO phase. These results provide compelling evidence of the successful coexistence and structural integration of ZnO and MIP-202(Zr) within the ZMIP nanocomposite. Furthermore, the corresponding elemental mapping images (Figure 5e) corroborate this observation, showing a uniform and well-distributed spatial dispersion of Zr, C, O, Cl, Zn, and N throughout the nanohybrid.

2.2. Optical Performance of the Synthesized Materials

As shown in Figure 6, the estimated E_g values of the green-synthesized ZnO nanoparticles, MIP-202(Zr) bio-MOF, and the ZMIP nanocomposite were estimated from three replicate measurements, yielding average values of 3.77 ± 0.04, 4.05 ± 0.06, and 2.74 ± 0.1 eV, respectively. The associated error bars represent the experimental uncertainty. The synthesized ZnO nanoparticles and MIP-202(Zr) bio-MOF both exhibited a white coloration, whereas the ZMIP nanocomposite displayed a slightly different off-white appearance. This observable color variation indicates the formation of the composite and suggests a subsequent alteration in its bandgap properties. The pronounced reduction in E_g upon integrating ZnO nanoparticles into the MIP-202(Zr) framework demonstrates the successful modification of its electronic structure, leading to improved visible-light absorption and, consequently, enhanced photocatalytic performance [34]. The decrease in bandgap energy observed for the ZMIP nanocomposite relative to its individual constituents (ZnO and MIP-202(Zr) bio-MOF) can be ascribed to intense interfacial coupling and synergistic electronic interactions between the semiconductor phase (ZnO) and the MOF. Upon incorporation of ZnO nanoparticles into the MIP-202(Zr) matrix, the formation of a heterojunction interface promotes partial overlap and alignment of the conduction and valence band edges of both materials, thereby narrowing the overall effective bandgap [21,35]. Moreover, electronic delocalization and charge redistribution occurring at the ZnO–MOF interface may generate intermediate electronic states or defect levels, which diminish the energy barrier for electron excitation. In addition, the organic linkers derived from L-aspartic acid can enhance π–π conjugation and facilitate localized electronic transitions, collectively contributing to the apparent reduction in bandgap energy and improved visible-light absorption [36]. Similar behavior has been observed in previous studies. For instance, Parsa et al. [37] demonstrated that depositing ZnO onto the MIL-100(Fe) MOF effectively reduced the bandgap from 2.82 eV for pristine MIL-100(Fe) to 2.18 eV for the composite. Likewise, Harisankar et al. [38] reported that the ZnO@MOF-5 composite showed a reduced band gap of 3.7 eV, compared with 3.8 eV for pristine MOF-5 and 3.44 eV for pure ZnO.

Additionally, bandgap estimation using Tauc plots can be significantly affected by the presence of long absorption tails at low energies, which render the extracted E_g values highly sensitive to the selected fitting region. These tails are often attributed to sub-bandgap absorption processes associated with structural defects, surface disorder, or dangling bonds that introduce localized electronic states within the bandgap. Such defect-related states can act as trap levels, lowering the effective excitation energy required for electron transitions [39,40]. Additionally, the observed behavior may also be linked to Urbach tails, which arise from lattice disorder and phonon interactions, leading to an exponential absorption edge [41]. In the case of ZnO–MOF composites, the generation of defect states at the ZnO/MIP-202 interface, along with possible electronic perturbations induced by the organic linkers, could further contribute to the extended absorption tail. While these features complicate the precise determination of E_g, they also indicate enhanced light-harvesting ability, as the defect-mediated states facilitate absorption in the visible region. Thus, the presence of absorption tails, although challenging for bandgap fitting, may play a beneficial role in improving photocatalytic activity by broadening the spectral response of the catalyst [42,43].

Based on the estimated electronegativities of the MIP-202(Zr) bio-MOF (C₂₄H₃₄N₆O₃₂Zr₆) and ZnO nanoparticles, which were 6.5 and 5.74 eV, respectively, the conduction band potentials (E_CB) were calculated as –0.03 eV for the MIP-202(Zr) MOF and –0.65 eV for ZnO. The corresponding valence band potentials (E_VB) were determined to be 4.02 eV and 3.12 eV, respectively. These findings reveal the distinct electronic structures of the two materials, suggesting that their combination can promote interfacial charge transfer within the composite system.

2.3. Contribution of Multiple Processes

Figure 7 illustrates the degradation efficiency of CBZ across different systems at pH 7, catalyst dosage of 1 g/L, initial CBZ concentration of 20 mg/L, and a reaction time of 60 min. The adsorption efficiency toward CBZ was first assessed under dark conditions. The green-synthesized ZnO nanoparticles, MIP-202(Zr) bio-MOF, and the ZMIP nanocomposite exhibited adsorption efficiencies of 50.09, 24.48, and 56.82%, respectively, confirming the superior adsorption efficiency of the composite material toward CBZ, resulting from its active surface sites and well-developed porous structure (Figure 5a,b). Under visible light irradiation, direct photolysis of CBZ was negligible, with only 7.26% removal in the light-only system. In contrast, the ZnO/light and MIP-202(Zr)/light systems achieved 65.18 and 30.38% CBZ degradation, respectively, whereas the ZMIP/light system reached 76.59%. In the presence of visible light, photoexcited electrons and holes are generated, which subsequently participate in redox reactions to produce ROS such as ^•OH and O₂^•− radicals [11]. These ROS serve as key agents in the degradation of CBZ molecules, acting synergistically with the adsorption capacity of the green-synthesized materials. The ZMIP nanocomposite exhibited markedly higher photocatalytic activity under light illumination compared to the individual ZnO nanoparticles and MIP-202(Zr) bio-MOF, demonstrating the synergistic effect of combining ZnO nanoparticles with the MIP-202(Zr) bio-MOF in enhancing CBZ removal. This improvement is primarily ascribed to its reduced bandgap energy (Figure 6), which enhances visible-light absorption and thereby promotes the generation of a greater quantity of reactive species [44].

2.4. Model Generation, Optimization, and Validation

The relationship between the independent variables and the CBZ degradation efficiency in the ZMIP/light system was mathematically modeled in the Minitab^® 22 statistical software and is described by the polynomial response function presented in Equation (1).

Y (%) = 95.2 − 0.277 D + 9.39 F − 4.42 G − 7.2 H + 0.00322 D² − 0.7596 F² − 0.0316 G² + 7.5 H² + 0.0148 DF + 0.0171 DG − 0.286 DH − 0.0069 FG − 0.98 FH + 2.84 GH

(1)

where D represents the reaction time (min), F is the solution pH, G denotes the initial CBZ concentration (mg/L), and H indicates the photocatalyst dosage (g/L).

Table 1. Experimental design parameters along with the corresponding actual and predicted CBZ removal efficiencies.

Run	Coded Values of Parameters				Actual Values of Parameters				CBZ Removal (%)
	D	F	G	H	D	F	G	H	Actual	Predicted
1	−1	0	−1	0	30	7	15	1	86.87	82.43
2	−1	−1	0	0	30	3	20	1	63.01	66.84
3	0	0	0	0	60	7	20	1	76.58	76.49
4	−1	1	0	0	30	11	20	1	49.23	51.49
5	0	1	−1	0	60	11	15	1	65.74	67.12
6	−1	0	0	1	30	7	20	1.25	83.17	84.08
7	1	1	0	0	90	11	20	1	72.65	71.18
8	0	1	0	1	60	11	20	1.25	70.00	68.07
9	0	0	0	0	60	7	20	1	76.54	76.49
10	0	1	1	0	60	11	25	1	49.34	48.18
11	0	1	0	−1	60	11	20	0.75	48.90	49.74
12	1	0	−1	0	90	7	15	1	93.12	93.44
13	0	−1	−1	0	60	3	15	1	76.73	78.64
14	1	−1	0	0	90	3	20	1	79.32	79.43
15	0	−1	0	1	60	3	20	1.25	85.77	81.83
16	1	0	0	−1	90	7	20	0.75	80.08	79.93
17	0	0	0	0	60	7	20	1	76.59	76.49
18	0	0	1	1	60	7	25	1.25	77.07	80.53
19	0	0	0	0	60	7	20	1	76.32	76.49
20	1	0	0	1	90	7	20	1.25	96.25	95.93
21	0	0	−1	1	60	7	15	1.25	90.18	92.10
22	0	−1	0	−1	60	3	20	0.75	60.75	59.58
23	−1	0	1	0	30	7	25	1	62.05	58.63
24	0	−1	1	0	60	3	25	1	60.88	60.25
25	−1	0	0	−1	30	7	20	0.75	58.43	59.50
26	0	0	1	−1	60	7	25	0.75	52.68	53.14
27	1	0	1	0	90	7	25	1	78.56	79.90
28	0	0	−1	−1	60	7	15	0.75	80.00	78.91
29	0	0	0	0	60	7	20	1	76.41	76.49

provides the full design matrix, incorporating both actual and coded levels of the four independent variables considered in this study. It also lists the corresponding experimental CBZ removal efficiencies together with the values predicted by the quadratic response surface model. A total of 29 experimental runs were conducted under different photocatalytic conditions within the ZMIP/light system to establish the reliability and predictive accuracy of the model. The experimentally measured CBZ removal efficiencies exhibited considerable variation, ranging from 48.9 to 96.25%, thereby covering a broad operational window and ensuring a comprehensive evaluation of the system performance. The close correspondence between experimental and predicted values further highlights the adequacy of the developed model in describing the degradation process.

The optimal operating conditions were determined using the response optimizer in the Minitab^® 22 statistical software based on maximizing the CBZ removal efficiency as a target. The resulting optimum values were an initial CBZ concentration of 15 mg/L, a photocatalyst dosage of 1.25 g/L, pH 6, and a reaction time of 90 min, as depicted in

Table 2. Derived optimum levels of the independent variables and CBZ removal efficiency under the optimum conditions.

Parameter	Optimum Value
Reaction time (min)	90
pH	6
Initial CBZ concentration (mg/L)	15
Photocatalyst dose (g/L)	1.25
CBZ removal efficiency (Laboratory)	99.37%
CBZ removal efficiency (Model)	98.84%

. To validate the reliability of the quadratic polynomial model, a confirmatory experiment was performed under these conditions, yielding a removal efficiency of 99.37% (Figure 7). This closely matched the predicted value of 98.84%, with a minimal deviation of 0.53%. The strong agreement between experimental data and predicted values demonstrates the robustness of the quadratic polynomial model in predicting CBZ photodegradation within the ZMIP/light system and its effectiveness in optimizing operational variables.

In addition, the degradation efficiency of CBZ in the ZMIP/light system was evaluated against the performance of a range of photocatalytic systems incorporating different metal oxide–MOF composite photocatalysts for the elimination of various pharmaceutical organic contaminants, as presented in

Table 3. Photocatalytic efficiencies of diverse metal oxide–MOF hybrid photocatalysts for the degradation of various pharmaceutical contaminants.

Photocatalyst	Light Source	Pollutant	Operating Conditions	Removal Ratio (%)	Reference
Bi4O7/MIL-68(In)-NH2	Visible light: Xenon lamp (300 W)	CBZ	[Pollutant]o = 50 mg/L, [Catalyst]o = 1 g/L, and reaction time = 120 min.	92.7	[45]
TiO2/MIL-101(Cr)	UV-A LED (20 W)	CBZ	[Pollutant]o = 12 mg/L, [Catalyst]o = 2 g/L, and reaction time = 60 min.	99	[46]
ZnO/ZIF-9	Ultraviolet light	Tetracycline	[Pollutant]o = 20 mg/L, [Catalyst]o = 0.25 g/L, and reaction time = 60 min.	87.7	[47]
ZnO/UiO-66/NH2	Visible light: Xenon lamp (50 W, 0.01–105 HZ)	Tetracycline	[Pollutant]o = 20 mg/L, [Catalyst]o = 0.1 g/L, and reaction time = 90 min.	61	[48]
ZnO/MIL-53(Al)	Visible light: Metal-halide lamp (Venture, 400 W)	Trimethoprim	[Pollutant]o = 10 mg/L, pH = 7, flowrate = 5 mL/min, and reaction time = 240 min.	93.5	[49]
ZnO-CuO/TMU-5	Visible light: Metal halide lamp (400 W, λ = 510 nm)	Tetracycline	[Pollutant]o = 30 mg/L, [Catalyst]o = 1 g/L, and reaction time = 180 min.	62	[50]
GO-WO3/UiO-66	Visible light: Mercury lamp (400 W)	Tetracycline	[Pollutant]o = 20 mg/L, [Catalyst]o = 1.67 g/L, pH = 7, and reaction time = 70 min.	84	[51]
TiO2/MIL-101(Cr)	Visible light: Xenon lamp (300 W, λ > 400 nm)	Tetracycline	[Pollutant]o = 10 mg/L, [Catalyst]o = 0.2 g/L, and reaction time = 45 min.	94.85	[52]
TiO2/UiO-66-NH2	Visible light: Xenon lamp (350 W)	Tetracycline	[Pollutant]o = 20 mg/L, [Catalyst]o = 0.12 mg TCN/mg catalyst, and reaction time = 60 min.	75	[53]
Bi2WO6/MIL-88B(Fe)	Visible light: Xenon lamp (500 W)	Tetracycline	[Pollutant]o = 10 mg/L, [Catalyst]o = 50 mg/L, and reaction time = 90 min.	96.4	[54]
Bi2WO6/MIL-125 (Ti)-NH2	Visible light: Xenon lamp (300 W)	Tetracycline	[Pollutant]o = 20 mg/L, [Catalyst]o = 0.4 g/L, and reaction time = 120 min.	69.19	[55]
ZnO/MIP-202(Zr)	Visible light: Metal halide lamp (400 W, λ = 510 nm)	CBZ	[Pollutant]o = 15 mg/L, [Catalyst]o = 1.25 g/L, pH = 6, and reaction time = 90 min.	99.37	This study

. The results highlight the superior efficacy of the proposed system.

2.5. Analysis of Variance

ANOVA was carried out using the Minitab^® 22 statistical software to evaluate the adequacy and reliability of the developed response surface model, as explained in

Table 4. ANOVA of the quadratic model for CBZ degradation response.

Source	DF	Sum of Squares	Mean Square	F-Value	p-Value
Model	14	4760.41	340.03	46.18	0
Linear	4	3481.67	870.42	118.2	0
D-Reaction time (min)	1	787.64	787.64	106.96	0
F-pH	1	415.36	415.36	56.41	0
G-CBZ concentration (mg/L)	1	1046.45	1046.45	142.11	0
H-Catalyst dose (g/L)	1	1232.21	1232.21	167.33	0
Square	4	1167.02	291.76	39.62	0
D (min) × D (min)	1	54.53	54.53	7.4	0.017
F × F	1	958.05	958.05	130.1	0
G (mg/L) × G (mg/L)	1	4.06	4.06	0.55	0.47
H (g/L) × H (g/L)	1	1.41	1.41	0.19	0.668
2-Way Interaction	6	111.71	18.62	2.53	0.072
D (min) × F	1	12.64	12.64	1.72	0.211
D (min) × G (mg/L)	1	26.32	26.32	3.57	0.08
D (min) × H (g/L)	1	18.36	18.36	2.49	0.137
F × G (mg/L)	1	0.08	0.08	0.01	0.921
F × H (g/L)	1	3.84	3.84	0.52	0.482
G (mg/L) × H (g/L)	1	50.48	50.48	6.86	0.02
Error	14	103.09	7.36	–	–
Lack-of-Fit	10	103.04	10.3	737.56	0
Pure Error	4	0.06	0.01	–	–
Total	28	4863.5	–	–	–

R² = 97.88%, R² (adjusted) = 95.76%.

. Being very close to unity, the R² (97.88%) and adjusted R² (95.76%) highlight the robustness and reliability of the regression, confirming that the model can accurately describe the experimental data. The close agreement between the two coefficients, with only a 2.12% discrepancy, indicates strong consistency between the predicted and experimental responses, thereby confirming the statistical validation of the fitted polynomial model in estimating the influence of experimental variables on CBZ removal efficiency in the ZMIP/light system. In addition, the high R² value indicates that 97.88% of the variation in CBZ removal efficiency can be explained by the model, whereas just 2.12% of the variation cannot be accounted for in the developed equation. The statistical adequacy and significance of the model was further verified by the high F-value (46.18) and the very low p-value (< 0.05). This is supported by the minimal difference observed between the predicted and experimental responses (Table 2). Moreover, several model terms (D, F, G, H, DF, DG, DH, FG, FH, GH, D², F², G² and H²) were identified as statistically significant, as their p-values were less than 0.05.

2.6. Influence of Operational Factors

The F-value was employed to assess the relative influence of each variable on the response. As presented in the ANOVA results (Table 4), the photocatalyst dose exhibited the highest F-value (167.33), considerably exceeding those of the other parameters. Accordingly, the magnitude of the variables’ effects on CBZ removal efficiency within the ZMIP/light system can be ranked as follows, photocatalyst dose > initial CBZ concentration > reaction time > solution pH, highlighting the dominant role of photocatalyst dosage in governing the degradation process.

To provide a clear visualization of how the interaction among the independent operational parameters (photocatalyst dosage, initial CBZ concentration, solution pH, and reaction time) affects CBZ degradation in the ZMIP/light system, contour plots were generated using the Minitab^® 22 statistical software, with two variables represented as coordinate axes, as shown in Figure 8.

The effect of ZMIP dosage on CBZ degradation efficiency was evaluated within the range of 0.75–1.25 g/L, as explained in Table 1 and

Table 5. Factors and design levels for the photocatalytic degradation of CBZ in the ZMIP/light system.

Independent Parameter	Code	Unit	Levels
			−1	0	1
Reaction time	D	min	30	60	90
pH	F	–	3	7	11
Initial CBZ concentration	G	mg/L	15	20	25
Photocatalyst dose	H	g/L	0.75	1	1.25

. The combined effects of reaction time and photocatalyst dosage on CBZ removal, under fixed conditions of pH 7 and an initial CBZ concentration of 20 mg/L, are illustrated in Figure 8a. The results demonstrated that CBZ removal efficiency increased proportionally with both photocatalyst dosage and reaction time. The highest degradation efficiency (96.25%) was achieved at a catalyst dosage of 1.25 g/L after 90 min. The improved CBZ degradation at the optimum dosage is attributed to the increased number of active sites on the photocatalyst surface, which enhances photon absorption and promotes higher ROS generation, thereby accelerating pollutant removal [56]. However, exceeding the optimal photocatalyst loading resulted in reduced degradation efficiency. This decline is primarily attributed to light scattering and turbidity at higher catalyst concentrations, which hinder effective light penetration and limit photon access to active sites, thereby suppressing charge carrier excitation [57]. In addition, particle agglomeration at higher dosages can further reduce the effective surface area and overall catalytic performance [58].

The influence of initial CBZ concentration on its degradation efficiency was investigated in the range of 15–25 mg/L at pH 7 and a ZMIP dosage of 1 g/L (Table 1 and Table 5). The contour plot showing the combined effects of CBZ concentration and reaction time is presented in Figure 8b. The results demonstrated a decline in removal efficiency with increasing initial CBZ concentration. The maximum degradation efficiency (93.12%) was achieved at 15 mg/L after 90 min. At low CBZ concentrations, the generated ROS are sufficient to degrade pollutants efficiently. However, as CBZ concentration increases, the available ROS become inadequate to sustain complete degradation, thereby reducing efficiency and requiring longer reaction times to reach comparable performance [59]. In contrast, since the number of active sites on the catalyst surface remains fixed, excess CBZ molecules (above the optimum value of 15 mg/L) can saturate these sites rapidly and suppress ROS generation, thereby hindering the degradation process [60]. In addition, at higher CBZ concentrations, pollutant molecules can absorb incident photons, thereby reducing the light available for electron–hole pair generation and reducing the CBZ removal efficiency [61].

The impact of solution pH on CBZ degradation efficiency was investigated in the range of 3–11 at a ZMIP dosage of 1 g/L and an initial CBZ concentration of 20 mg/L (Table 1 and Table 5). The combined effects of pH and reaction time on CBZ removal are presented in Figure 8c. Since the pKa of CBZ is approximately 13.9, it remains essentially neutral across the tested pH range (3–11) [62]. Therefore, its adsorption onto the ZMIP surface is not driven by electrostatic interactions but rather by π–π stacking, hydrophobic interactions, and hydrogen bonding [63]. On the other hand, the pHpzc of the ZMIP photocatalyst was determined to be 8.6 (Figure S4), indicating a positively charged surface at pH < 8.6 and a negatively charged surface at pH > 8.6 [64]. The photocatalytic performance followed the order pH 3 > pH 7 > pH 11, with the highest degradation efficiency (79.32%) obtained at pH 3 after 90 min. At pH 3, the ZMIP surface is positively charged while CBZ remains neutral, enabling interactions through hydrogen bonding (via the amide group) and π–π stacking [65]. In addition, the positively charged surface enhances hydroxyl ion (OH⁻) adsorption near the active sites of the photocatalyst, where they are oxidized by photogenerated holes to form ^•OH radicals, thus improving degradation efficiency under acidic conditions [66]. At neutral pH 7, the catalyst surface remains positively charged, but less strongly than at pH 3. CBZ also remains neutral, allowing adsorption through π–π stacking and hydrogen bonding, though to a lesser extent than under acidic conditions. As a result of the lower proton concentration, ^•OH generation at pH 7 is only moderate, leading to a lower CBZ degradation efficiency compared with pH 3. In contrast, at pH 11, the catalyst surface becomes strongly negatively charged (pH > pHpzc), while CBZ remains neutral. The negative surface charge repels OH⁻ ions from the catalyst surface, thereby reducing their availability for oxidation by photogenerated h⁺ to produce ^•OH radicals. In addition, the negative surface potential hinders electron transfer from the conduction band to electron acceptors in solution, leading to electron accumulation within the catalyst or at surface states. This accumulation promotes electron–hole recombination and suppresses ROS generation [67]. Additionally, weaker hydrogen bonding and diminished surface–substrate interactions under alkaline conditions further reduce photocatalytic efficiency [68].

As shown in Figure 8, extending the irradiation time from 30 to 90 min significantly enhanced CBZ degradation efficiency. This enhancement is due to the prolonged exposure of photocatalyst active sites to light, which allows greater photon absorption. The longer irradiation also increases the generation of reactive species and photogenerated carriers. As a result, the probability of interactions between ROS and CBZ molecules becomes higher [69]. However, this positive effect persists only up to a certain duration and extended irradiation beyond the optimum time can decrease the overall degradation efficiency. Prolonged illumination periods may promote the recombination of photogenerated electron–hole pairs and reduce the delocalization of charge carriers, leading to an equilibrium between radical generation and recombination [70].

2.7. Photocatalytic Degradation Kinetics of CBZ

The degradation kinetics of CBZ in the ZMIP/light system were analyzed using the pseudo-first-order and pseudo-second-order kinetic models. The experiments were conducted at neutral pH (7), with a catalyst dosage of 1 g/L, and a reaction period of 60 min. Three initial CBZ concentrations, 15, 20, and 25 mg/L, were examined under these experimental conditions. As shown in Figure 9a, the apparent first-order rate constants (K₁) for CBZ degradation were 0.0327 (R² = 0.9936) min⁻¹, 0.0254 (R² = 0.9935), and 0.022 (R² = 0.9828) min⁻¹ at initial CBZ concentrations of 15, 20, and 25 mg/L, respectively. In addition, the apparent second-order rate constants (K₂) were 0.0069 (R² = 0.9792) min⁻¹, 0.0036 (R² = 0.9645), and 0.0025 (R² = 0.9365) min⁻¹, respectively, at the same CBZ concentrations, as presented in Figure 9b. The higher R² obtained from the first-order model relative to the second-order model demonstrates that the photodegradation of CBZ in the ZMIP/light system is best fitted by first-order kinetics. The pseudo-first-order kinetic model is widely regarded as the most suitable framework for describing the photocatalytic degradation behavior of organic pollutants in aqueous environments [71]. Moreover, the results confirmed a decline in the CBZ degradation rate with increasing initial CBZ concentration. These findings are consistent with the previously discussed results regarding the effect of initial pollutant concentration on the system performance. On the other hand, the slight deviations from linearity observed in the kinetic plots likely arise from complex surface adsorption–desorption equilibria and the simultaneous occurrence of photogenerated electron–hole recombination during the photocatalytic reaction. These processes influence the apparent degradation rate and are not fully captured by simplified kinetic models that assume uniform active site distribution and ideal reaction conditions. Nevertheless, the pseudo-first-order kinetic model remains a widely accepted and effective empirical approach for describing and comparing the photocatalytic degradation behavior of organic contaminants in heterogeneous systems [72,73].

2.8. The Reusability of the ZMIP Photocatalyst

The reusability of the ZMIP photocatalyst was assessed under the optimum conditions (pH 6, catalyst dosage of 1.25 g/L, initial CBZ concentration of 15 mg/L, and reaction duration of 90 min) by recovering and reutilizing the ZMIP composite powder after each cycle over five consecutive photocatalytic runs. Following each photodegradation cycle, the ZMIP powder was collected and subjected to repeated centrifugation (EBA 20, Hettich, Beverly, MA, USA) with ultrapure water and ethanol until the supernatant was entirely transparent. The recovered material was then dried at 60 °C for 24 h in an electric oven (ED 53, Binder, Tuttlingen, Germany) to eliminate residual moisture. The dried photocatalyst was subsequently introduced into a newly prepared TCN solution to commence the next experimental run [74,75]. As illustrated in Figure 10a, the nanocomposite demonstrated only a slight decline in CBZ removal efficiency, achieving 99.37, 96.80, and 94.54% after the first, second, and third successive cycles, respectively. At the end of the fourth and fifth cycles, the degradation efficiency decreased to 85.23 and 77.97%, respectively, indicating that prolonged reaction times might be necessary in the later stages to achieve the desired pollutant degradation. The gradual decline in the catalytic performance of the ZMIP photocatalyst over successive cycles can be ascribed to multiple factors, including partial catalyst loss during recovery, diminished surface area due to potential nanoparticle agglomeration, leaching of Zn²⁺ ions into the reaction medium, and progressive surface fouling caused by the adsorption of CBZ molecules and their degradation intermediates onto the active sites [76,77]. Despite these challenges, the ZMIP composite retained sufficient photocatalytic activity, confirming sustained reactive species generation under continuous light exposure and highlighting its potential for long-term and practical environmental applications.

2.9. Total Organic Carbon (TOC) Mineralization in the ZMIP/Light System

The efficiency of TOC removal during the photocatalytic degradation of CBZ using the ZMIP photocatalyst was evaluated under the optimum conditions as an additional indicator of the treatment performance. At an initial CBZ concentration of 15 mg/L and an initial TOC concentration of 6.24 mg/L, the system achieved a TOC mineralization efficiency of 84.39% after 90 min of irradiation, as presented in Figure 10b. These results confirm the promising potential of the ZMIP photocatalyst for application in real industrial wastewater treatment. Although TOC removal progressively increased over time during the photocatalytic reaction, the overall TOC mineralization efficiency was lower than the CBZ removal efficiency under identical conditions (Figure 7), owing to the sequential formation of organic intermediates derived from CBZ [78]. The presence of these organic intermediates extends the persistence of organic load in the reaction medium and competes with the parent pollutant for the generated ROS, thereby incomplete TOC mineralization was attained [79]. Complete mineralization of CBZ to environmentally benign end products (CO₂ and H₂O) requires prolonged reaction times to enable repeated attacks by ROS on the intermediate degradation products [80].

2.10. Photocatalytic Degradation Mechanism of CBZ

Considering the calculated E_CB and V_CB potentials of the synthesized ZnO nanoparticles and MIP-202(Zr) bio-MOF, a plausible charge transfer pathway within the ZMIP/light photocatalytic system is proposed. Upon illumination, photons excite the ZMIP composite, promoting electrons from the VB to the conduction band CB and generating corresponding holes in the VB. This process creates e⁻/h⁺ pairs within both components of the composite, initiating charge separation [81]. Subsequently, the photogenerated electrons in the CB of ZnO migrate to the CB of the MIP-202(Zr) MOF, while the holes in the VB of MIP-202(Zr) are transferred to the VB of ZnO, facilitating effective charge separation across the ZMIP heterojunction [79]. The interfacial charge transfer between the green-synthesized ZnO nanoparticles and MIP-202(Zr) bio-MOF effectively suppresses the recombination of photogenerated e⁻/h⁺ pairs, thereby prolonging the lifetime of charge carriers and enhancing the photocatalytic degradation efficiency of the target pollutant [82]. Moreover, the photogenerated electrons can react with dissolved molecular oxygen to form O₂^{• −} radicals, while the holes in the VB could oxidize H₂O molecules and/or OH⁻ ions to generate hydroxyl radicals (^•OH) [83]. Subsequently, the generated O₂^{• −} radicals can react with protons (H⁺) to produce H₂O₂, which may further interact with electrons to yield additional ^•OH radicals [84]. The photogenerated holes, O₂^{• −} radicals, and ^•OH radicals can interact with organic molecules adsorbed near the active sites on the ZMIP photocatalyst surface, thereby enhancing the oxidative degradation of the targeted organic contaminant [85]. Figure 11 provides a schematic representation of the charge transfer pathway between MIP-202(Zr) framework and ZnO nanoparticles, along with the proposed mechanism for ROS generation resulting from their interaction.

To quantify the participation of the generated ROS in the photocatalytic reaction within the ZMIP/light system, ammonium oxalate, benzoquinone, and tert-butyl alcohol were employed to deactivate the oxidation ability of h⁺, O₂^{• −} radicals, and ^•OH radicals, respectively [86,87]. The experiments were employed under the optimum conditions (pH 6, catalyst dosage of 1.25 g/L, initial CBZ concentration of 15 mg/L, and reaction duration of 90 min). As demonstrated in Figure 12a, the degradation efficiency of CBZ was reduced to 82.9, 68.65, and 50.15% upon the addition of 1 mM of ammonium oxalate, benzoquinone, and tert-butyl alcohol, respectively, compared to 99.37% without any quenching agents. These results demonstrated that all types of ROS contributed to CBZ photocatalytic decomposition process, confirming the oxidation mechanism. However, the dominant reactive species in CBZ degradation within the ZMIP/light system followed the order ^•OH > O₂^{• −} > h⁺, indicating that ^•OH radicals were the most effective contributors, while photogenerated holes played the least significant role.

2.11. Degradation of Other Organic Pollutants

A systematic evaluation of the photocatalytic performance of the ZMIP/light system was carried out for the degradation of various pharmaceutical contaminants, including doxorubicin hydrochloride, tetracycline, paracetamol, and ibuprofen. Each compound was investigated separately under the operating conditions optimized for CBZ degradation (pollutant concentration = 15 mg/L, pH = 6, ZMIP dosage = 1 g/L, and reaction time = 90 min). The degradation efficiencies were 82.93% for doxorubicin hydrochloride, 76.84% for tetracycline, 72.08% for paracetamol, and 67.71% for ibuprofen, in comparison with 99.37% observed for CBZ, as presented in Figure 12b. These outcomes highlight the capability of the ZMIP photocatalyst to effectively degrade a broad spectrum of emerging pharmaceutical pollutants, demonstrating not only the potential of the ZMIP/light system in practical water purification applications but also the versatility of the operating conditions originally optimized for CBZ. The observed variations in degradation efficiencies among different compounds can be primarily attributed to the fact that the applied operating parameters were optimized for CBZ removal, emphasizing the necessity of customizing catalytic conditions to maximize pollutant-specific degradation. Furthermore, detailed investigation into the factors governing pollutant selectivity and the intrinsic limitations of ZMIP would provide valuable insights for improving the catalyst’s applicability in diverse water treatment contexts.

2.12. CBZ Proposed Degradation Paths

Based on the detected m/z peaks of CBZ intermediates formed during its degradation by the ZMIP photocatalyst under light irradiation (Figure S5), the corresponding by-products were identified, and the photocatalytic degradation pathways of CBZ were subsequently proposed, as illustrated in Figure 13. In Path A, the olefinic double bond within the heterocyclic ring of CBZ (m/z = 236) was initially attacked by superoxide and hydroxyl radicals, leading to its oxidation into the intermediate product P1 (m/z = 301) which then underwent intramolecular cyclization and dehydration to form the by-product P2 (m/z = 267) [88]. Through decarboxylation processes, the intermediate P3 (m/z = 224) was transformed into P4 (m/z = 196) [89]. Then, through successive oxidation steps involving photogenerated holes, iodate, hydroxyl, and superoxide radicals, the intermediate P4 was further transformed into the smaller aromatic P5 (m/z = 138) [90]. In Path B, CBZ underwent oxidative transformation initiated by superoxide radicals and the photogenerated holes, forming the by-product P6 (m/z = 253) [6]. The intermediate P6 was converted into P3 through ring contraction and cleavage reactions [91]. Similar to Path A, the by-product P3 may undergo further oxidation to produce the smaller degradation products P4 and P5. Alternatively, P3 can experience N–C bond cleavage induced by superoxide radicals, leading to the formation of P7 (m/z = 180), which can subsequently be oxidized to generate P4 and P5 [92]. In both proposed pathways, P5 was eventually mineralized into CO₂ and H₂O, highlighting the efficiency of the ZMIP/light system in facilitating the stepwise degradation of CBZ [46].

The MS data revealed that the peak at m/z = 267 (P2) exhibited the highest signal intensity among the detected intermediates, suggesting that the initial oxidation of the olefinic double bond, followed by intramolecular cyclization and dehydration, represents a major and kinetically favored transformation route. In contrast, the peak at m/z = 301 (P1), attributed to a hydroxylated intermediate, appeared with significantly lower intensity, indicating that this species is transient and rapidly converted to P2. The prominence of P2 (m/z = 267) following the parent compound CBZ (m/z = 236) confirms its substantial accumulation during degradation. Meanwhile, intermediates associated with Path B (e.g., P6, m/z = 253) were detected with noticeably lower relative intensities, implying that this route proceeds only to a limited extent. Based on the relative abundance and persistence of the intermediates, particularly those corresponding to m/z = 267, it can be concluded that Path A is the predominant degradation route under the experimental conditions. This pathway ultimately leads to the formation of smaller aromatic intermediates, which are further mineralized into CO₂ and H₂O.

2.13. Effect of Coexisting Substances

Real wastewater is characterized by a complex composition of inorganic ions and organic compounds, which may hinder the photocatalytic degradation of organic contaminants [93]. To better simulate practical application scenarios, various real water samples collected from different natural sources were utilized as reaction solutions for the degradation of CBZ within the ZMIP/light system. The evaluated matrices included deionized water, tap water, seawater, lake water, and drain water. The collection sites of the raw wastewater samples are presented in Table S2, whereas their corresponding physicochemical characteristics are summarized in Table S3. In addition, a detailed description of pre-processing and conditioning procedures of these samples is provided in Text S1. All experiments were performed under the optimum conditions: pH 6, catalyst dosage of 1.25 g/L, and initial CBZ concentration of 15 mg/L. As illustrated in Figure 14a, the deionized water yielded the highest CBZ degradation efficiency among all tested water matrices, achieving 99.37% after 90 min. Tap water also demonstrated a relatively high removal efficiency of 92.81%, showing only a 6.6% reduction compared to deionized water. In contrast, significantly lower degradation efficiencies were observed in seawater (70.86%), lake water (54.06%), and drain water (31.08%). The higher CBZ removal efficiencies observed in deionized and tap water matrices can be attributed to their relatively low matrix complexity, as evidenced by their low turbidity and minimal concentrations of dissolved organic matter and total dissolved inorganic ions [94]. In contrast, the reduced photocatalytic performance of the ZMIP photocatalyst in sea, lake, and drain water is likely due to the increased complexity of these matrices, which may interfere with light penetration and active site availability, thereby inhibiting the overall degradation efficiency [95,96]. Natural organic matter (NOM) can adversely affect photocatalytic degradation by competing with the target pollutant for ROS, thereby diminishing its removal efficiency [97]. Additionally, the high aromaticity of NOM, reflected by elevated specific ultraviolet absorbance (SUVA) values (≥2 L/mg.m), promotes significant light absorption. This competitive absorption diminishes the photons available to excite the photocatalyst and the target pollutant, thereby limiting irradiation at the active sites and ultimately reducing ROS generation [98]. Moreover, NOM may adsorb onto the surface of the photocatalyst, obstructing active sites and further hindering the production of ROS essential for pollutant degradation. On the other hand, the elevated turbidity in lake, drain, and seawater matrices may scatter incident light, thereby limiting the penetration of light to the catalyst’s active sites [99]. In addition, the presence of inorganic salts in these matrices can inhibit the photocatalytic process by acting as scavengers of ROS, leading to the formation of less reactive radicals and ultimately reducing the degradation efficiency of CBZ [100].

Additionally, the effects of common coexisting substances such as NOM (e.g., humic acid), inorganic anions (Cl⁻, HCO₃⁻, NO₃⁻, SO₄²⁻), and inorganic cations (Mg⁺², Ca⁺², Na⁺, Cu⁺²) on the removal efficiency of CBZ in the ZMIP/light system were investigated. The initial CBZ concentration was set at 15 mg/L, with the solution pH maintained at 6, while the photocatalyst was introduced at a dosage of 1.25 g/L. After 60 min of reaction, the presence of humic acid at a concentration of 5 mg/L slightly enhanced the photocatalytic degradation of CBZ, increasing the removal efficiency to 81.5% compared to 76.59% in the absence of humic acid, as illustrated in Figure 14b. This suggests that the presence of a low level of humic acid can positively influence the photocatalytic process. Indeed, humic acid is a naturally occurring photosensitizer capable of absorbing light energy, thereby facilitating photochemical energy conversion and potentially promoting the generation of reactive species during the reaction [101]. However, a progressive inhibitory effect on CBZ degradation was observed with increasing concentrations of humic acid. Specifically, the removal efficiency declined to 67.84 and 59.23% at humic acid concentrations of 10 and 20 mg/L, respectively. This reduction is likely attributed to competitive interactions between humic acid and CBZ for ROS, which hampers the availability of ROS for CBZ oxidation [102]. These findings highlight the importance of removing excess humic substances, particularly at concentrations exceeding 5 mg/L, through a pre-treatment step when applying the proposed photocatalytic system for real wastewater treatment applications.

As shown in Figure 15a, the addition of 10 mM of Cl⁻, HCO₃⁻, NO₃⁻, and SO₄²⁻ resulted in a decrease in CBZ removal efficiency to 93.02, 85.76, 80.92, and 75.34%, respectively, after 90 min, compared to 99.37% in the absence of added anions. The results indicated that the Cl⁻ ions exhibited a negligible impact, while both HCO₃⁻ and NO₃⁻ significantly suppressed the photocatalytic activity. The most pronounced inhibition was observed with SO₄²⁻, indicating that the presence of sulfate in practical applications may substantially reduce the system’s degradation efficiency. Overall, the inhibitory effects of the tested anions on CBZ degradation within the ZMIP/light system followed the order: SO₄²⁻ > NO₃⁻ > HCO₃⁻ > Cl⁻. On the other hand, the addition of 2 mM of Mg⁺², Ca⁺², Na⁺, and Cu⁺² reduced the CBZ removal efficiency to 96.02, 87.59, 82.39, and 70.30%, respectively, after 90 min of reaction, as presented in Figure 15b. The addition of Mg⁺² had a negligible effect on CBZ removal, whereas the presence of Ca⁺² and Na⁺ led to a significant decline in the photocatalytic efficiency of the ZMIP/light system. The Cu²⁺ ions exhibited the strongest inhibitory effect. These findings suggest that the photocatalytic performance of the proposed system may be significantly reduced in real wastewater containing Ca⁺², Na⁺, and particularly Cu²⁺, while Mg⁺² is unlikely to cause substantial interference. The decline in degradation efficiency in the presence of inorganic ions can be explained by their dual interaction with ROS. On one hand, these ions may react with ROS to generate secondary radicals, which generally exhibit lower oxidation potential and reactivity [103]. On the other hand, they can also act as quenchers that directly consume ROS, thereby reducing their overall availability. Both pathways ultimately diminish the oxidative capacity of the system [104]. Moreover, competition between inorganic ions and CBZ molecules for the limited reactive sites on the photocatalyst surface may further suppress the photocatalytic degradation performance [105].

2.14. Remediation of Real Industrial Discharge

Real pharmaceutical wastewater was treated using the ZMPI photocatalyst in the presence of visible light to assess the system’s practical photocatalytic performance. Due to the expected complex composition of the water sample, the reduction in TOC was selected as a representative indicator of overall treatment efficiency. The photocatalytic process was conducted over five successive cycles, each with a reaction time of 90 min, at pH 6 and ZMIP dosage of 1.25 g/L. As illustrated in Figure 16a, the TOC mineralization efficiencies of the real sample (initial TOC = 75.34 mg/L) declined from 78.37% in the first cycle to 72.24, 63.76, 50.50, and 38% after the repeated cycles, respectively. The results demonstrated a substantial reduction in TOC values, indicating effective degradation of the organic content. However, minimal TOC removal was observed during the initial stages of the reaction, likely due to the formation of intermediate by-products that retain organic carbon and compete with the parent organic compounds for ROS [106]. In the later stages of the reaction, further degradation of organic intermediates occurred, leading to a notable decrease in TOC levels [107]. However, complete mineralization requires the generation of sufficient ROS to fully degrade both the original organic pollutants and the intermediate by-products [97]. These findings underscore the promising potential of the ZMPI composite for the effective treatment of industrial wastewater.

2.15. Effect of Introducing Highly Reactive Inorganic Oxidizing Agents

The recombination of photogenerated e⁻/h⁺ pairs is a major drawback in photocatalytic systems, as it results in significant energy loss and decreases the availability of charge carriers required for redox reactions [108]. To overcome this limitation, reactive inorganic oxidants such as H₂O₂, KHSO₅, K₂S₂O₈, and KIO₄ were introduced into the reaction solution before initiating the photocatalytic reaction. These oxidants function as efficient electron acceptors by capturing photogenerated electrons on the photocatalyst surface. This process suppresses the recombination of e⁻/h⁺ pairs, preserving more charge carriers for redox reactions. In addition, the presence of these oxidants facilitates the production of more ROS, which play a decisive role in accelerating the photocatalytic degradation process [109]. The experiments were carried out at pH 7, using a ZMIP dosage of 1 g/L and an initial CBZ concentration of 20 mg/L. Each oxidant was introduced separately into the reaction solution at a concentration of 3 mM. Following the addition of these components, the light lamp was switched on to initiate the photocatalytic reaction. The results, shown in Figure 16b, demonstrate that the addition of KIO₄ produced the most pronounced improvement in CBZ degradation efficiency (98.93%), followed by K₂S₂O₈ (92.15%), KHSO₅ (89.22%), and H₂O₂ (81.86%). In contrast, the system without any added oxidant achieved a lower degradation efficiency of 76.59%.

The superior CBZ removal efficiency observed with KIO₄, relative to other oxidants, is likely due to the comparatively extended I–O bond length in periodate ions (IO₄⁻) (approximately 1.78 Å), which enhances its susceptibility to activation and subsequently promotes the formation of various ROS [110]. Periodate ions possess a strong electron-accepting capability, enabling them to effectively capture photogenerated electrons on the surface of the ZMIP photocatalyst, as illustrated in Equation (2) [111]. This electron-scavenging process suppresses e⁻/h⁺ recombination, thereby extending the lifetime of photogenerated holes and enhancing their availability for subsequent oxidative reactions. Moreover, upon visible-light illumination, IO₄⁻ ions can produce iodate radicals (IO₃^•), as illustrated in Equation (3) [112]. In addition, IO₄⁻ plays a key role in the formation of O₂^{• −} radicals and singlet oxygen (¹O₂), as described in Equations (4) and (5) [113,114]. On the other hand, the oxygen-containing functional groups on the ZMIP nanocomposite surface, as indicated in the FTIR spectrum (Figure 2), can interact with IO₄⁻ ions to produce additional IO₃^• radicals as well as ^•OH radicals, as demonstrated in Equations (6) and (7) [74]. These ROS actively participate in the oxidative degradation of CBZ, breaking it down into smaller intermediate compounds, CO₂, and H₂O [34].

The addition of K₂S₂O₈ significantly enhanced the degradation efficiency of CBZ, primarily due to the ability of persulfate ions (S₂O₈²⁻) to scavenge photogenerated electrons from the illuminated ZMIP photocatalyst, thereby suppressing e⁻/h⁺ recombination and facilitating the formation of sulfate radicals (SO₄^{• −}) as potent ROS, as illustrated in Equation (8) [115]. Furthermore, the homolytic cleavage of the peroxide bond in S₂O₈²⁻ under light irradiation leads to the generation of additional SO₄^{• −} (Equation (9)) [116]. Hydroxyl radicals can also be produced via the reaction of SO₄^{• −} with H₂O molecules or OH⁻ ions, as shown in Equations (10) and (11) [117,118]. In addition, other ROS, such as O₂^{• −} radicals and ¹O₂, may be formed through subsequent secondary reactions, as presented in Equations (12)–(14) [119,120]. Moreover, the oxygen-containing functional groups present on the surface of the photocatalyst can react with S₂O₈²⁻ ions, leading to the generation of additional SO₄^{• −} radicals, as illustrated in Equations (15) and (16) [75]. The slightly lower CBZ degradation efficiency with K₂S₂O₈ compared to KIO₄ may result from the quenching effect of S₂O₈²⁻ ions on SO₄^{• −} and ^•OH radicals. This interaction converts the highly reactive radicals into the less active species (S₂O₈^{• −}), as illustrated in Equations (17) and (18) [77].

Analogous to K₂S₂O₈, the incorporation of KHSO₅ into the ZMIP/light photocatalytic system enhanced CBZ degradation efficiency, primarily due to the dual role of peroxymonosulfate anion (HSO₅⁻) in accepting photogenerated electrons and concurrently producing SO₄^{• −}, as illustrated in Equation (19) [121]. In addition, HSO₅⁻ can undergo direct photolysis under light irradiation, resulting in the formation of both SO₄^{• −} and ^•OH radicals, as shown in Equation (20) [122]. Further, O₂^{• −} radicals can be generated through the reaction between HSO₅⁻ ions and SO₄^{• −} radicals, as presented in Equations (21) and (22) [123,124]. Like the K₂S₂O₈-based system, ^•OH radicals and the ¹O₂ species can be generated through the reactions presented in Equations (10) and (11) and (Equation (14)), respectively. The functional groups on the surface of the ZMIP catalyst can also activate HSO₅⁻ to produce SO₄^{• −} radicals, as explained in Equations (23) and (24) [125,126].

Nevertheless, the CBZ removal efficiency achieved with KHSO₅ was lower than that obtained with K₂S₂O₈. This difference can be explained by the structural characteristics of the oxidants. The O–O bond length in the HSO₅⁻ ions (1.453 Å) is shorter than that in S₂O₈²⁻ ions (1.497 Å). As a result, S₂O₈²⁻ ions are more easily activated and more efficient in producing ROS [96].

Although H₂O₂ possesses the capability to scavenge photogenerated electrons and can be activated under visible light to produce ^•OH radicals (Equations (25) and (26)) [122,123], its addition resulted in the lowest CBZ removal efficiency among the tested inorganic oxidants. This limited performance can be attributed to the relatively strong O–O bond in H₂O₂ molecules, which impedes its cleavage and activation by the photocatalyst, thereby restricting the generation of ROS [127]. Moreover, the lower oxidation potential and shorter half-life of ^•OH radicals, dominant in the H₂O₂-based system, further constrain its oxidative effectiveness when compared to SO₄^{• −} radicals, which are efficiently generated in the K₂S₂O₈ and KHSO₅ systems [75].

$${\mathrm{I}\mathrm{O}}_4^{-}+8{\mathrm{e}}^{-}+8{\mathrm{H}}^{+}\to {\mathrm{H}}_2\mathrm{O}+{\mathrm{I}}^{-}$$

(2)

$${\mathrm{I}\mathrm{O}}_4^{-}+\mathrm{h}\mathrm{v}\to {\mathrm{O}}^{•}+{\mathrm{I}\mathrm{O}}_3^{•}$$

(3)

$${3\mathrm{I}\mathrm{O}}_4^{-}+2\mathrm{O}{\mathrm{H}}^{-}\to {\mathrm{H}}_2\mathrm{O}+{3\mathrm{I}\mathrm{O}}_3^{-}+{2\mathrm{O}}_2^{•-}$$

(4)

$${\mathrm{I}\mathrm{O}}_4^{-}+{2\mathrm{O}}_2^{•-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{I}\mathrm{O}}_3^{-}+2\mathrm{O}{\mathrm{H}}^{-}+2({}^1\mathrm{O}_2)$$

(5)

$$\mathrm{ZMIP}–\mathrm{OOH}+{\mathrm{I}\mathrm{O}}_4^{-}\to \mathrm{ZMIP}–{\mathrm{OO}}^{•}+{}^{•}\mathrm{OH}+{\mathrm{I}\mathrm{O}}_3^{•}$$

(6)

$$\mathrm{ZMIP}–\mathrm{OH}+{\mathrm{I}\mathrm{O}}_4^{-}\to \mathrm{ZMIP}–{\mathrm{O}}^{•}+{}^{•}\mathrm{OH}+{\mathrm{I}\mathrm{O}}_3^{•}$$

(7)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+{\mathrm{e}}^{-}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(8)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+\mathrm{h}\mathrm{v}\to {2\mathrm{S}\mathrm{O}}_4^{•-}$$

(9)

$${\mathrm{S}\mathrm{O}}_4^{•-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{\mathrm{H}}^{+}+{\mathrm{O}\mathrm{H}}^{•}$$

(10)

$${\mathrm{S}\mathrm{O}}_4^{•-}+\mathrm{O}{\mathrm{H}}^{-}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{\mathrm{O}\mathrm{H}}^{•}$$

(11)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{S}\mathrm{O}}_4^{-2}+3{\mathrm{H}}^{+}+{\mathrm{H}\mathrm{O}}_2^{-}$$

(12)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+{\mathrm{H}\mathrm{O}}_2^{-}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{\mathrm{H}}^{+}+{\mathrm{O}}_2^{•-}$$

(13)

$${\mathrm{O}}_2^{•-}+{\mathrm{O}\mathrm{H}}^{•}\to \mathrm{O}{\mathrm{H}}^{-}+{}^1\mathrm{O}_2$$

(14)

$$\mathrm{ZMIP}–\mathrm{OOH}+{{\mathrm{S}}_2\mathrm{O}}_8^{-2}\to \mathrm{ZMIP}–{\mathrm{OO}}^{•}+{\mathrm{H}\mathrm{S}\mathrm{O}}_4^{-}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(15)

$$\mathrm{ZMIP}–\mathrm{OH}+{{\mathrm{S}}_2\mathrm{O}}_8^{-2}\to \mathrm{ZMIP}–{\mathrm{O}}^{•}+{\mathrm{H}\mathrm{S}\mathrm{O}}_4^{-}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(16)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+{\mathrm{S}\mathrm{O}}_4^{•-}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{{\mathrm{S}}_2\mathrm{O}}_8^{•-}$$

(17)

$${{\mathrm{S}}_2\mathrm{O}}_8^{-2}+{\mathrm{O}\mathrm{H}}^{•}\to \mathrm{O}{\mathrm{H}}^{-}+{{\mathrm{S}}_2\mathrm{O}}_8^{•-}$$

(18)

$${\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+{\mathrm{e}}^{-}\to \mathrm{O}{\mathrm{H}}^{-}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(19)

$${\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+\mathrm{h}\mathrm{v}\to {\mathrm{O}\mathrm{H}}^{•}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(20)

$${\mathrm{S}\mathrm{O}}_4^{•-}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to {\mathrm{S}\mathrm{O}}_5^{•-}+{\mathrm{H}\mathrm{S}\mathrm{O}}_4^{-}$$

(21)

$${\mathrm{S}\mathrm{O}}_5^{•-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{S}\mathrm{O}}_4^{-2}+{\mathrm{H}}^{+}+{\mathrm{O}}_2^{•-}$$

(22)

$$\mathrm{ZMIP}–\mathrm{OOH}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to \mathrm{ZMIP}–{\mathrm{OO}}^{•}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(23)

$$\mathrm{ZMIP}–\mathrm{OH}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to \mathrm{ZMIP}–{\mathrm{O}}^{•}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{S}\mathrm{O}}_4^{•-}$$

(24)

$${{\mathrm{H}}_2\mathrm{O}}_2+{\mathrm{e}}^{-}\to {\mathrm{O}\mathrm{H}}^{-}+{\mathrm{O}\mathrm{H}}^{•}$$

(25)

$${{\mathrm{H}}_2\mathrm{O}}_2+\mathrm{h}\mathrm{v}\to {2\mathrm{O}\mathrm{H}}^{•}$$

(26)

3. Materials and Methodology

3.1. Real Wastewater Samples

Table S2 lists the sampling locations of the raw wastewater specimens, while Table S3 tabulates their measured chemical and physical properties. Moreover, procedural details regarding wastewater sample preparation protocol are documented in Text S1.

3.2. Chemicals and Materials

Detailed information on the chemicals list and materials employed in this study is provided in Text S2 of the Supplementary Materials (SM).

3.3. Green Synthesis of the ZnO Nanoparticles

The comprehensive procedures for the preparation of water lettuce extract and the subsequent green synthesis of ZnO nanoparticles are described in Texts S3 and S4.

3.4. Construction of the MIP-202(Zr) Bio-MOF

The MIP-202(Zr) bio-MOF was synthesized via a mild hydrothermal method, as provided in Text S5 of the SM.

3.5. Fabrication of the ZMIP Nanocomposite

During the preparation of the MIP-202(Zr) bio-MOF, 5 g of the green-synthesized ZnO nanoparticles was introduced into the boiling flask immediately before initiating reflux heating, to achieve a 1:1 mass ratio of the two components. The subsequent synthesis steps were carried out in the same manner, and the resulting nanocomposite was designated as ZMIP.

3.6. Physicochemical Characterization of the Synthesized Materials

An outline of the analytical approaches employed to investigate the physicochemical characterizations of the fabricated catalysts is given in Text S6.

3.7. Point of Zero Charge

The point of zero charge (pHpzc) of the synthesized ZMIP nanocomposite was identified through the powder addition technique, commonly referred to as the pH drift method, as outlined in Text S7.

3.8. Optical Properties

The optical characteristics of the synthesized materials were comprehensively assessed following the procedures outlined in Text S8.

3.9. Experimental Set-Up and Degradation Study

An accurately weighed amount of CBZ was dissolved in deionized water to prepare a stock solution with a concentration of 50 mg/L. The photocatalytic degradation of CBZ was conducted in batch mode using a 250 mL glass reactor. Illumination was provided by a centrally mounted metal halide lamp (400 W) emitting visible light at λmax of 510 nm, with a 10 cm headspace left above the system. The reactor temperature was maintained at ambient value (25 ± 1 °C), and the solution pH was adjusted, as necessary, using 0.1 M solutions of sodium hydroxide (NaOH) and/or sulfuric acid (H₂SO₄). A digital magnetic stirrer hotplate (MSH-20D, DAIHAN Scientific Co., Ltd., Seoul, Republic of Korea) was used to agitate the reactor contents and regulate the reaction temperature. A multiparameter analyzer (HQ440d, HACH, Loveland, CO, USA) coupled with a pH sensor (PHC301, HACH, Loveland, CO, USA) was employed to measure and verify the solution pH. To initiate the photocatalytic reaction, the light source was switched on, and the pre-specified amount of the photocatalyst was added to 100 mL of pollutant solution. All experiments were conducted in triplicate to ensure the reproducibility of the results. A graphical overview of the experimental workflow is presented in Figure S6.

3.10. Experimental Design and Statistical Analysis

RSM was employed to assess the photocatalytic performance of the proposed system and to analyze the individual and interactive effects of the selected independent process variables on the dependent parameter or response (CBZ removal efficiency). Moreover, this methodology was utilized to optimize the independent variables based on maximizing the response as a target across three experimental levels coded as −1 (low), 0, and +1 (high). The independent parameters and their tested ranges were reaction time (D, 30–60 min), solution pH (F, 3–11), initial CBZ concentration (G, 15–25 mg/L), and photocatalyst dosage (H, 0.75–1.25 g/L), as presented in Table 5. The independent variables were incorporated into the Box–Behnken Design (BBD) for modeling. Including five replicated central points (Cp), a total of 29 experimental runs (Table 1) was determined using Equation (27) [128]. These experiments were employed to develop the design matrix and to predict the CBZ removal efficiency based on Equation (28) [129].

$$N=2\mathrm{F}\left(\mathrm{F}-1\right)+\mathrm{C}\mathrm{p}$$

(27)

$$\mathrm{Y}(\mathrm{\%})={\beta }_0+\sum _{i=1}^k{\mathsf{\beta }}_i X_{\mathrm{i}}+\sum _{i=1}^k{\mathsf{\beta }}_{ii} {\mathrm{X}}_i^2 \sum _{i=1}^{k-1}\sum _{j=i+1}^k{\mathsf{\beta }}_{ij} X_{\mathrm{i}\mathrm{j}} +e$$

(28)

where N represents the total number of experimental runs and F denotes the number of independent variables. The term Y (%) denotes the CBZ removal efficiency (response). Additionally, the β₀, β_i, β_ii, and β_ij terms represent the regression coefficients corresponding to the intercept, linear, quadratic, and interaction terms, respectively. Further, the coded independent variables are expressed as X_i and X_j (with i and j ranging from 1 to k), where k is the total number of independent variables (k = 4 in this study), and e corresponds to the model error (residual term). The experimental design was structured to provide a comprehensive assessment of variable interactions, thereby overcoming the inherent limitations of conventional one-factor-at-a-time (OFAT) approaches [130].

The Minitab^® 22 statistical software was employed to perform regression analysis and to construct a polynomial model describing the overall process, thereby elucidating the relationship between the independent variables and response. Analysis of variance (ANOVA) was applied to evaluate the adequacy and reliability of the developed model [131]. The model’s validity was determined based on the correlation coefficient (R²), Fisher’s statistical values (F-values), and p-values, where a high F-value coupled with a p-value ≤ 0.05 indicated strong statistical significance in describing the relationship between the independent variables and the response [132].

3.11. Analytical Methods

The analytical methodologies applied throughout this investigation are thoroughly described and systematically outlined in Text S9 and Table S4.

3.12. Kinetic Study

In this work, the degradation kinetics of CBZ in the proposed system were evaluated by applying both pseudo-first-order and pseudo-second-order kinetic models to identify the model that most accurately characterizes the CBZ degradation process, as presented in Text S10.

4. Conclusions

This work reports the first successful integration of ZnO nanoparticles with the bio-MOF MIP-202(Zr) to construct a composite photocatalyst capable of efficiently degrading CBZ in real pharmaceutical wastewater under visible-light irradiation. The characterization analyses verified the successful fabrication and interaction between Zn and the MIP-202(Zr) framework. Under identical conditions (pH = 7, catalyst dosage = 1 g/L, initial CBZ concentration = 20 mg/L, and reaction time = 60 min), the photocatalytic efficiency of the ZMIP nanocomposite reached 76.59%, which was significantly higher than that of pure ZnO and MIP-202(Zr) due to the reduction in bandgap in ZMIP to 2.74 ± 0.1 eV and the improvement of charge carrier separation. The optimum conditions were identified using RSM as a reaction time of 90 min, solution pH of 6, initial CBZ concentration of 15 mg/L, and a photocatalyst dosage of 1.25 g/L, under which CBZ removal achieved 99.37% with a corresponding TOC mineralization of 84.39%. The ZMIP photocatalyst exhibited high stability and reusability, maintaining CBZ degradation efficiencies of 99.37, 96.80, 94.54, 85.23, and 77.97% across five successive cycles, thereby confirming its sustained catalytic performance during repeated operation. Under visible-light irradiation, the ZMIP heterojunction promotes efficient charge separation by transferring electrons from ZnO to MIP-202(Zr) and holes in the opposite direction. This process suppresses e⁻/h⁺ recombination and facilitates the generation of ROS (^•OH and O₂^{• −} radicals), which synergistically drive CBZ oxidation. Further, the quenching experiments identified ^•OH radicals as the dominant reactive species in the CBZ degradation process, followed by O₂^{• −} radicals, while the photogenerated holes played the least significant role. On the other hand, the degradation pathways of CBZ were elucidated through the identification of transformation products using UHPLC-MS analysis. Furthermore, under the optimum operating conditions, the ZMIP/light photocatalytic system demonstrated strong catalytic activity toward a range of pharmaceutical contaminants (15 mg/L), achieving removal efficiencies of 82.93% for doxorubicin hydrochloride, 76.84% for tetracycline, 72.08% for paracetamol, and 67.71% for ibuprofen, in comparison with 99.37% for CBZ. These findings highlight the broad applicability of the ZMIP/light system, demonstrating its potential as a versatile photocatalyst for the effective removal of structurally diverse pharmaceutical pollutants. In addition, the system maintained high CBZ degradation performance across different water matrices, with efficiencies of 92.81% in tap water, 70.86% in seawater, 54.06% in lake water, and 31.08% in drain water, compared to 99.37% in deionized water. Further, the degradation system exhibited lower performance in the presence of high NOM levels and inorganic salts; therefore, these compounds should be removed or minimized before applying the ZMIP/light system in real wastewater treatment to achieve optimal photocatalytic performance. Significantly, the ZMIP/light system demonstrated consistent and robust performance in the treatment of real industrial wastewater obtained from a pharmaceutical plant. The substantial TOC removal achieved (78.37% after 90 min) highlights the practical applicability of the ZMIP/light system, emphasizing its potential as a reliable photocatalytic technology for the advanced treatment of complex industrial wastewater streams. In addition, the introduction of highly reactive inorganic oxidants markedly enhanced CBZ degradation in the ZMIP/light system. Among the oxidants tested, KIO₄ exhibited the greatest enhancement, achieving 98.93% removal, followed by K₂S₂O₈ (92.15%), KHSO₅ (89.22%), and H₂O₂ (81.86%), compared with 76.59% in the absence of additives. This enhancement indicates that integrating the ZMIP/light system with suitable inorganic oxidants, particularly KIO₄, can significantly improve photocatalytic efficiency, providing a promising approach for the near-complete elimination of pharmaceutical pollutants. In the future, the economic feasibility and environmental impacts of the proposed degradation system can be investigated. Furthermore, to advance toward practical implementation, the system should be evaluated using real wastewater matrices, alongside the design and optimization of large-scale photocatalytic reactors.