Abstract
Environmental pollution has increased significantly because of the extensive use of antibiotics, such as penicillins, which end up in ecological water through direct disposal and wastewater effluents. Most of these antibiotics exist in trace concentrations in the environment, but they can cause detrimental effects to living species. The negative side effects include carcinogenesis, mutagenesis, allergy reactions, teratogenesis and the development of antibiotic‐resistant bacterial strains. Therefore, it is necessary to develop effective materials as adsorbents in solid-phase extraction (SPE) to remove and preconcentrate antibiotics from the water matrix before detection and quantification by an analytical instrument. Metal-organic frameworks (MOFs) have emerged as promising adsorbents due to their large surface area porosity and excellent adsorption capacity.
This study explored UiO-66(Zr) and MIL-53(Al) adsorbents in dispersive micro-solid phase extraction (D-μ-SPE) of amoxicillin (AMX), penicillin g (PNG), piperacillin (PIP) and penicillin v (PNV) from water. UiO-66(Zr) was prepared from waste polyethylene terephthalate (PET) as a source of ligand, whereas MIL-53(Al) was prepared from commercial terephthalic acid as a ligand precursor. The functional groups, thermal stability, crystallinity, textural properties, and morphology of MOFs were characterized using various techniques, including X-ray diffraction spectroscopy (XRD), scanning electron microscopy (SEM), Fourier transform infrared (FTIR) spectroscopy, transmission electron microscopy (TEM), thermogravimetric analysis (TGA), and surface area and porosity analyser.
The antibiotics were separated and analysed by high-performance liquid chromatography with diode array detection (HPLC-DAD) after the D-μ-SPE step. The performance of the MOF-D-μ-SPE technique for penicillins was optimized to determine the effect of experimental variables such as sample pH, mass of adsorbent (MA), extraction time (ET), eluent volume (EV) and desorption time (DT) were investigated using fractional factorial design (FrFD) and Box-Behnken design (BBD) approach. The optimum conditions were utilized to investigate adsorption affinity, capacity, and mechanisms. The adsorption isotherm for PET-derived UiO-66(Zr) was best fitted to the Freundlich isotherm model, confirming that the adsorption process was heterogeneously and multilayered. On the other hand, MIL-53(Al) fitted the Langmuir isotherm model, suggesting homogeneous and monolayered adsorption. MOFs adsorption data fitted the pseudo-second-order kinetics models.
vi
The PET-derived UiO-66(Zr) maximum adsorption capacities for AMX and PNG were 139 and 147 mg/g, respectively. The maximum adsorption capacities of MIL-53(Al) for PIP and PNV were 135 mg/g and 133 mg/g, respectively. The effectiveness of the developed D-μ-SPE methods was confirmed by its remarkable analytical performances, such as low limits of detection (LOD) (0.19-0.21 μg/L) and linear range of 0.6-1000 μg/L for PET-derived UiO-66(Zr). MIL-53(Al) produced low LODs of 0.10-0.18 μg/L and a linear range of 0.3-800 μg/L. The feasibility and applicability of the validated D-μ-SPE/HPLC-DAD method were confirmed by applying it to real water samples. Finally, the spike recoveries in wastewater ranged from 89.9-101%, while the precision (intraday and interday) of the method was less than 5%. AMX and PNG were detected and quantified in wastewater influent and effluent in concentrations of 1.51-4.71 μg/L and 0.195-1.19 μg/L, respectively. PIP and PNV were found in the range of 0.71–1.12 μg/L and 0.32–0.45 μg/L in wastewater influent and effluent, respectively.