Assessing Zn12O12 nanocage for the elimination of ciprofloxacin from drinking water using DFT simulations Authors: Qaisar Ali, Hamad Khan, Salman Khan, Mahtab Alam, Ajmal Shah, Abdullah F. AlAsmari, Fawaz Alasmari, Momin Khan Published: 2026 ABSTRACT According to the biochemical theory, the presence of pollutants such as antibiotics is a worldwide issue since they foster antibiotic resistance. A detailed DFT analysis was carried out to examine the adsorption capabilities of the Zn12O12 nanocage for ciprofloxacin (CIP). Five optimal structures of Zn12O12–CIP complexes were identified, designated as CMP-1, CMP-2, CMP-3, CMP-4, and CMP-5. CMP-3 shows the most favorable adsorption with Eads values of 48.99 kcal/mol, representing the strongest interaction without any structural deformation of the nanocage. The interaction occurs through the bonding of the F and O atoms of CIP with the Zn and O atoms of the nanocage, respectively. FMO indicated a reduction in the bandgap for the Zn12O12-CIP complexes, making the complexes more reactive and suggesting the favorable adsorption characteristics of the nanocage for the drug. The NBO analysis revealed that the CMP-3 complex had the highest charge transfer of 0.139 e with the bond distances of 1.92 Å and 1.25 Å, respectively. The presence of covalent and weak electrostatic interactions between the nanocage and CIP molecules was confirmed by QTAIM, NCI, and RDG analyses. The thermodynamic assessment showed that the adsorption process is spontaneous. The extremely high adsorption energy values limit the complete desorption process and reduce the regeneration ability of the Zn12O12 nanocage. Overall, the nanocage has proven theoretically to be a potential candidate for environmental remediation applications.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-57419-z. Introduction Antibiotics are widely used for the treatment of a broad spectrum of infections, including urinary tract, respiratory, and central nervous system infections1. Ciprofloxacin (CIP) is a broad-spectrum antibiotic belonging to the fluoroquinolone class, which is one of the most commonly used antibiotics in the treatment of various bacterial infections2. During CIP use, a significant portion of CIP is excreted in urine, and thus the drug continuously enters water bodies due to incomplete absorption in the human body3. Other major sources of CIP contamination in water include discharges from hospital effluents, pharmaceutical manufacturing, and domestic sewage4,5. The persistence of CIP in water bodies has become a serious environmental and public health concern because prolonged exposure induces bacterial resistance, endocrine disruption, neurological complications such as seizures, glycemic responses, gastrointestinal disorders, and tendon disorders6. In addition, CIP has high aqueous solubility over a wide pH range and possesses a chemically stable molecular structure, making it resistant to biodegradation and difficult to remove from water systems. Therefore, the efficient removal of CIP from wastewater is an urgent environmental challenge7. Different treatment methods have been investigated for CIP remediation, such as biological degradation8, photolysis9, membrane separation10, ion-exchange processes11, and chemical oxidation methods12. Despite their advantages, these methods are frequently limited by negative factors such as high operation costs, low selectivity, and low efficiency that prevent the application of such methods on a large scale13–15. However, over the last few years, nanomaterials have become promising candidates in the context of environmental applications due to their unique physicochemical properties16–21. Different nanocages, such as B12N12, Al12N12, and Be12O12 have previously been explored for the adsorption and sensing of pharmaceutical pollutants owing to their remarkable structural stability and adsorption performance. For instance, B12N12 nanocages have been reported as efficient adsorbents for CIP removal due to favorable adsorption interactions22, while Al12N12 and Be12O12 nanocages have demonstrated strong adsorption behavior toward pharmaceutical contaminants in previous studies23–31. Zn12O12 nanocage is particularly attractive due to its high chemical stability, semiconducting nature, biocompatibility, and strong surface reactivity, making it a suitable candidate for adsorption and sensing applications. The Zn12O12 nanocage may provide a considerable charge transfer interaction and improved adsorption behavior due to the presence of active (ZnO) sites. Therefore, exploring the adsorption mechanism of CIP on Zn12O12 nanocage is essential for evaluating their potential in wastewater remediation, as previously reported for CS2 adsorption32–40. In this work, the density functional theory (DFT) calculations were employed to investigate the adsorption behavior of CIP on Zn12O12 nanocage. This study provides a comprehensive theoretical assessment of CIP adsorption on zinc oxide adsorbent through detailed electronic, structural, and thermodynamic analyses. Frontier molecular orbital (FMO), natural bond orbital (NBO), charge transfer assessment, and dipole moment analyses were performed to understand the adsorption mechanism. Moreover, non-covalent interaction (NCI) and reduced density gradient (RDG) were utilized to explain the nature of interactions. The effect of solvent on the adsorption mechanism was studied using the polarizable continuum model (PCM), whereas thermodynamics and regeneration studies were carried out to evaluate the efficiency and reusability of Zn12O12 nanocages for the adsorption of CIP from wastewater. Results and discussions Geometries optimization of Zn12O12 nanocage, CIP, and complexes The optimized structure of the Zn12O12 nanocage consisted of six four-membered rings (4MRs) and eight six-membered rings (6MRs), as shown in Fig. 1. The analysis indicated two types of bonds in the nanocage: Zn-O bonds in both 4MRs and 6MRs, and Zn-Zn bonds in the 4MRs, with bond lengths of about 1.90 Å and 1.80 Å, respectively. The longer Zn–O bond length in the 4MRs compared to the 6MRs is due to higher angular strain in the smaller ring system, which is consistent with previous reports. The bond angle analysis revealed that the O-Zn-O and Zn-O-Zn angles in 6MRs and 4MRs are 126.2°, 92.2°, 112.0°, and 86.4°, respectively. These data are consistent with the earlier documented information, which validates the geometrical characteristics of the nanocage41,42. The electronic properties of the nanocage are indicated by the molecular electrostatic potential (MEP) map in Fig. 2. The yellow and red regions around the oxygen atoms correspond to negative electrostatic potential and represent electron-rich regions, whereas the blue regions surrounding the Zn atoms indicate positive electrostatic potential and electron-deficient regions. These regions are favorable for interactions with complementary sites of the ciprofloxacin molecule. The MEP distribution suggests that the O atoms of Zn12O12 may interact with the hydrogen atom of the hydroxyl group in CIP, whereas the Zn atoms may interact with the oxygen atoms of the carbonyl groups. In addition, possible interaction between Zn atoms and the fluorine atom of CIP was also considered during complex construction. Molecular complexes were formed using these predicted interaction sites, which were subsequently analyzed. To investigate the adsorption characteristics of CIP on the Zn12O12 nanocage, five different configurations (CMP-1 to CMP-5) were optimized using the same computational approach. These geometries are depicted in Fig. 3, and the results are summarized in Table 1. The most stable interactions were observed when H and F atoms of CIP and O atoms of the C=O functional groups interact with Zn and O atoms in the nanocage. The greatest adsorption energies were observed for CMP-3, in which the carbonyl O atom and hydroxyl H atom of CIP bonded to Zn and O atoms of the Zn12O12 nanocage. The adsorption energies for all five configurations were more negative than − 11 kcal/mol, suggesting dominant chemisorption character in the adsorption process. CMP-3 had the most negative adsorption energy of about − 48.99 kcal/mol, demonstrating the most stable geometry among the optimized complexes. The bond lengths of the complexes indicate significant interactions: CMP-1 has bond length values of 1.99 Å and 2.22 Å, and CMP-2 has a bond length value of about 1.98 Å. Similarly, the bond distance values of the two interacting positions in CMP-3 are about 1.92 Å and 1.25 Å, respectively. Likewise, the bond distances were found to be 2.18 Å and 2.07 Å in CMP-4 and CMP-5, correspondingly. It is interesting to note that the smallest bond lengths were observed in CMP-3 (1.25 Å in H-O and 1.92 Å in O-Zn), which supports the findings that CMP-3 is the most stable geometry with the highest adsorption energy values among the optimized complexes. Additionally, the bond elongation study demonstrated significant interaction in the complexes. The two C = O bonds were elongated in CMP-1 by 0.02 Å and 0.002 Å, respectively, from their original bond lengths of 1.99 Å and 2.22 Å. In the case of CMP-2 and CMP-3, 0.03 Å and 0.04 Å bond elongations in the C=O group were observed. Similarly, CMP-4 and CMP-5 showed a similar elongation of C-F and C-O bonds, with a bond elongation of 0.03 Å and 0.04 Å, respectively, indicating a significant interaction between the adsorbent and the adsorbate. The hydrogen bond in CMP-3 is elongated by 0.2 Å, indicating the strongest interaction among the optimized configurations, suggesting the highest stability of this structure. The cumulative bond elongation values of C=O and H-O were 0.04 Å and 0.2 Å, respectively, supporting the comparatively greater stability of CMP-3 among the investigated complexes. Table 1Adsorption energy (Eads) values, HOMO-LUMO (FMO), and charge transfer (NBO) analyses.SystemDFT-D36-31G(d, p) Eads (kcal/mol)DFT-D3SPK-DZPEads (kcal/mol)DFT-D3SPK-TZPEads (kcal/mol)ELUMO (eV)EHOMO (eV)Eg (eV)Q (e)Zn12O12–––− 2.85− 6.944.09–Ciprofloxacin–––− 1.21− 5.704.49–CMP-1− 39.37− 26.54− 25.84− 2.29− 5.683.39− 0.09CMP-2− 39.32− 30.17− 27.90− 2.04− 6.114.07− 0.10CMP-3− 48.99− 43.56− 39.05− 2.51− 5.833.32− 0.14CMP-4− 23.60− 15.52− 13.26− 2.47− 6.143.67− 0.003CMP-5− 33.48− 26.16− 23.62− 2.67− 5.933.26− 0.06 Table 1Adsorption energy (Eads) values, HOMO-LUMO (FMO), and charge transfer (NBO) analyses.SystemDFT-D36-31G(d, p) Eads (kcal/mol)DFT-D3SPK-DZPEads (kcal/mol)DFT-D3SPK-TZPEads (kcal/mol)ELUMO (eV)EHOMO (eV)Eg (eV)Q (e)Zn12O12–––− 2.85− 6.944.09–Ciprofloxacin–––− 1.21− 5.704.49–CMP-1− 39.37− 26.54− 25.84− 2.29− 5.683.39− 0.09CMP-2− 39.32− 30.17− 27.90− 2.04− 6.114.07− 0.10CMP-3− 48.99− 43.56− 39.05− 2.51− 5.833.32− 0.14CMP-4− 23.60− 15.52− 13.26− 2.47− 6.143.67− 0.003CMP-5− 33.48− 26.16− 23.62− 2.67− 5.933.26− 0.06 Adsorption energy (Eads) values, HOMO-LUMO (FMO), and charge transfer (NBO) analyses. Fig. 1Optimized geometries of Zn12O12 nanocage and CIP. Optimized geometries of Zn12O12 nanocage and CIP. Fig. 2Molecular electrostatic potential (MEP) map of Zn12O12 nanocage and CIP. Molecular electrostatic potential (MEP) map of Zn12O12 nanocage and CIP. Fig. 3Optimized complexes of Zn12O12 nanocage with CIP. Optimized complexes of Zn12O12 nanocage with CIP. The accuracy of the adsorption energy values was further checked using higher basis sets. The adsorption energies obtained at the DFT-D3 level with SPK-DZP and SPK-TZP basis sets depicted Eads values greater than − 11 kcal/mol, depicting chemisorption and supporting the stability trends observed for all complexes. At the SPK-DZP basis set, the adsorption energies range from − 15.52 kcal/mol to -43.56 kcal/mol, while at the SPK-TZP basis set, they vary from − 13.26 kcal/mol to -39.05 kcal/mol, as given in Table 1, suggesting that all complexes exhibited chemisorption across different basis sets. Global reactivity descriptors The global reactivity descriptors provide important insights into the chemical stability, charge transfer ability, and reactivity of the adsorbent before and after the adsorption of adsorbate. The Zn12O12 nanocage possesses the highest chemical hardness (η) 2.05 eV and the lowest softness (S) 0.24 eV⁻¹, indicating relatively high electronic stability and lower polarizability. In contrast, the free adsorbate exhibits slightly higher hardness, η = 2.25 eV, reflecting its comparatively lower reactivity. Upon complexation, all complexes show a noticeable decrease in hardness (1.63–2.03 eV) along with an increase in softness (0.25–0.31 eV⁻¹) mentioned in Table 2, suggesting considerable charge transfer between the interacting species and improved electronic responsiveness. Among all complexes, CMP-5 shows the lowest hardness of 1.63 eV, and the highest softness of 0.31 eV− 1, suggesting comparatively higher polarizability and stronger electronic interaction relative to the other complexes. Similarly, CMP-3 also demonstrates high reactivity with 1.67 eV hardness and 0.03 eV− 1 softness. The electronegativity values increase after complexation compared with isolated adsorbate, indicating greater electron-attracting capability of the complexes. The electronegativity index significantly increases for the adsorption systems, particularly for CMP-5 (5.67 eV) and CMP-3 (5.23 eV), demonstrating electronic accepting character and stabilization through charge transfer. The negative chemical potential values for all systems suggest thermodynamic stability. Overall, the global reactivity descriptors suggest interaction between the CIP adsorbate and the Zn12O12 adsorbent. Table 2Global reactivity descriptors such as ionization potential (IP), electron affinity (EA), electronegativity (χ), chemical hardness (η), chemical softness (S), chemical potential (µ), and global electrophilicity index (ω).SystemEHOMO (eV)ELUMO (eV)Eg (eV)IP (eV)EA (eV)χ (eV)η (eV)S (eV⁻¹)µ (eV)ω (eV)Zn12O12− 2.85− 6.944.096.942.854.902.050.24− 4.905.85Ciprofloxacin− 1.21− 5.704.495.701.213.462.250.22− 3.462.66CMP-1− 2.29− 5.693.395.692.293.991.700.29− 3.994.68CMP-2− 2.05− 6.114.076.112.054.082.030.25− 4.084.10CMP-3− 2.51− 5.843.325.842.514.181.670.30− 4.185.23CMP-4− 2.47− 6.143.676.142.474.311.840.27− 4.315.05CMP-5− 2.67− 5.933.265.932.674.301.630.31− 4.305.67 Table 2Global reactivity descriptors such as ionization potential (IP), electron affinity (EA), electronegativity (χ), chemical hardness (η), chemical softness (S), chemical potential (µ), and global electrophilicity index (ω).SystemEHOMO (eV)ELUMO (eV)Eg (eV)IP (eV)EA (eV)χ (eV)η (eV)S (eV⁻¹)µ (eV)ω (eV)Zn12O12− 2.85− 6.944.096.942.854.902.050.24− 4.905.85Ciprofloxacin− 1.21− 5.704.495.701.213.462.250.22− 3.462.66CMP-1− 2.29− 5.693.395.692.293.991.700.29− 3.994.68CMP-2− 2.05− 6.114.076.112.054.082.030.25− 4.084.10CMP-3− 2.51− 5.843.325.842.514.181.670.30− 4.185.23CMP-4− 2.47− 6.143.676.142.474.311.840.27− 4.315.05CMP-5− 2.67− 5.933.265.932.674.301.630.31− 4.305.67 Global reactivity descriptors such as ionization potential (IP), electron affinity (EA), electronegativity (χ), chemical hardness (η), chemical softness (S), chemical potential (µ), and global electrophilicity index (ω). Frontier molecular orbitals (FMO) analysis and density of state (DOS) analysis The electronic response of the Zn12O12 nanocage toward CIP adsorption was further investigated using frontier molecular orbital (FMO) and density of states (DOS) analyses. The HOMO and LUMO plots are shown in Fig. 4 together with their respective DOS spectra in Fig. 5 of the free Zn12O12 nanocage and its complexes with adsorbate. It is noteworthy that upon complex formation, a reduction in HOMO energy levels (EHOMO) was observed, as well as changes in the energy bandgap, as presented in Table 1. The CMP-5 and CMP-3 configurations had the smallest energy gaps of 3.26 eV and 3.32 eV, respectively, as illustrated in Fig. 5. These lower bandgap values suggest comparatively greater orbital interaction and electronic perturbation upon adsorption. Conversely, CMP-2 exhibited the largest energy gap, which is approximately 4.07 eV, indicating weak interaction in CMP-2. The HOMO-LUMO gap value tends to decrease in general when the conduction electrons are more available, thus improving the electrical conductivity of the system. The increased electrical conductivity facilitates the generation of an electrical signal, which may contribute to an enhanced electronic response of the nanocage upon CIP adsorption. In Fig. 4, the FMO orbitals can be visualized with the oxygen atoms of the Zn12O12 nanocage being localized in the LUMO orbitals, whereas Zn atoms of the nanocage are surrounded by the HOMO orbitals. Upon complexation, a major redistribution of orbitals was observed. The LUMO orbitals are localized on the nanocage, and the HOMO shifts towards the adsorbate molecule in the resulting complexes, indicating the existence of intermolecular interactions between the two entities. Specifically, the diagrams visualized in Fig. 4 indicate the localization of the LUMO orbitals at the interacting atoms in CMP-1, CMP-2, and CMP-3. The orbital distributions are highly consistent with the Eads values, suggesting effective orbital overlap and significant interaction between the Zn12O12 nanocage and the adsorbate. Although the observed electronic perturbations indicate favorable adsorption-induced changes in the nanocage electronic structure, these theoretical results alone are insufficient to directly establish practical sensing or water-remediation performance under experimental conditions. Fig. 4Frontier molecular orbitals (FMO) and HOMO-LUMO analyses of the complexes of Zn12O12 nanocage with CIP. Frontier molecular orbitals (FMO) and HOMO-LUMO analyses of the complexes of Zn12O12 nanocage with CIP. Fig. 5Density of states (DOS) analysis of the complexes of Zn12O12 nanocage with CIP. Density of states (DOS) analysis of the complexes of Zn12O12 nanocage with CIP. Reduced density gradient (RDG) analysis and non-covalent interaction (NCI) analysis Reduced density gradient (RDG) analysis was used to determine the nature and strength of interactions between the CIP molecule and the Zn12O12 nanocage. Fig. 6 shows the plots of RDG against sign (λ2)ρ, as well as non-covalent interaction (NCI) isosurfaces. The technique provides important knowledge of both strong and weak interactions that take place in molecular complexes. In the NCI isosurfaces, green regions generally correspond to electrostatic interactions, blue regions are associated with hydrogen and halogen bonding, whereas red regions indicate steric repulsion. The existence of electrostatic interactions is shown by the green encircled spikes in the RDG plot in all five CIP-Zn12O12 complexes. The blue encircled spike in CMP-3 and CMP-4 indicates hydrogen and halogen bonding between the hydrogen atom of the hydroxyl group of the CIP and the oxygen atom of the nanocage and between a zinc atom of Zn12O12 and the fluorine atom of the adsorbate. Similarly, the red encircled spikes in the RDG plots suggest the existence of steric repulsion. Additionally, electrostatic interaction is present between the C = O oxygen atom of the CIP and Zn atoms of the Zn12O12 nanocage of CMP-1, CMP-2, and CMP-3, represented by green arrows in the NCI plots. Similarly, the green encircled spike in the RDG plot of CMP-5 and the green spot indicated by the green color arrow in the NCI plot show the presence of an electrostatic interaction. Also, there are some green patches at different sites of the CIP molecule, which indicate the possibility of intramolecular van der Waals forces. These results replicated the outcomes of the FMO and DOS analyses and suggest the existence of interactions between the Zn12O12 nanocage and the CIP drug. Fig. 6Reduced density gradient (RDG) and non-covalent interaction (NCI) analyses of the complexes of Zn12O12 nanocage with CIP. Reduced density gradient (RDG) and non-covalent interaction (NCI) analyses of the complexes of Zn12O12 nanocage with CIP. Quantum theory of atoms in molecules (QTAIM) analysis To further investigate the nature of intermolecular interactions between the Zn12O12 nanocage and the CIP molecule, the quantum theory of atoms in molecules (QTAIM) analysis was performed using the Gaussian 16 software suite (https://seagrid.org/). Fig. 7 gives the topological results in graphical form. QTAIM enables characterization of covalent and electrostatic interactions between the interacting species, which examines several electron density properties at bond critical points (BCPs), including the electron density (ρb), Laplacian of the electron density (∇2ρb), kinetic electron energy density (Gb), potential energy density (Vb), and total energy density (Hb). Types of interaction are classified according to the sign of ∇2ρb and Hb, and the values of the -Gb/Vb ratio. Negative Hb values are generally associated with shared-electron interactions, whereas positive Hb values are commonly associated with closed-shell interactions. The nature of the interactions between monomers in Zn12O12-CIP complexes was analyzed at BCPs 57, 29, 69, 72, and 83, with negative Hb values indicating shared interactions in the complexes, as indicated in Table 3. However, the Hb values at BCP 55 in CMP-1 were positive, suggesting predominantly electrostatic interaction at that position, which is further witnessed by the values of the -Gb/Vb ratio being more than unity. Also, the positive values of ∇²ρb and negative values of Hb at BCPs 57, 29, 69, 72, and 83 suggest the presence of partial interaction comprising both weak electrostatic and covalent interactions, which is further supported by the values of -Gb/Vb ratio falling in the range of 0.5 to 1.0. The BCP 51 in CMP-3 is consistent with relatively strong hydrogen-bonding interaction possessing covalent character, as indicated by the -Gb/Vb ratio being less than 0.5. The highest electron-density (ρb) values at BCP 69 also indicate that CMP-3 has the strongest interaction among all complexes, consistent with the highest Eads values. Similarly, high values of ρb were observed for CMP-2 at BCP 29, followed by CMP-1 at BCP 57. However, the Eads values of CMP-1 are higher, which can be explained by the fact that two BCPs (57 and 55) contribute to it. On the other hand, BCP 72 has the smallest values of ρb among the optimized complexes, suggesting comparatively weaker interaction strength for CMP-4 among the investigated configurations. These findings are in close agreement with the calculated Eads values of the complexes. Table 3Quantum theory of atoms in molecules (QTAIM) analysis.System Zn12O12BCPDensity of all electrons (ρb)laplacian (∇2ρb)Kinetic electron density GbPotential energy density VbEnergy density Hb− Gb/VbCMP-1570.0710.3360.105− 0.126− 0.0210.833550.0380.2020.050− 0.0490.000691.014CMP-2290.0740.3320.106− 0.130− 0.0230.820CMP-3690.0940.4010.137− 0.174− 0.0370.788510.147− 0.1820.090− 0.225− 0.1350.399CMP-4720.0380.2080.053− 0.054− 0.00120.978CMP-5830.0600.2690.081− 0.095− 0.0140.855 Table 3Quantum theory of atoms in molecules (QTAIM) analysis.System Zn12O12BCPDensity of all electrons (ρb)laplacian (∇2ρb)Kinetic electron density GbPotential energy density VbEnergy density Hb− Gb/VbCMP-1570.0710.3360.105− 0.126− 0.0210.833550.0380.2020.050− 0.0490.000691.014CMP-2290.0740.3320.106− 0.130− 0.0230.820CMP-3690.0940.4010.137− 0.174− 0.0370.788510.147− 0.1820.090− 0.225− 0.1350.399CMP-4720.0380.2080.053− 0.054− 0.00120.978CMP-5830.0600.2690.081− 0.095− 0.0140.855 Quantum theory of atoms in molecules (QTAIM) analysis. Fig. 7Quantum theory of atoms in molecules (QTAIM) analysis of the complexes of Zn12O12 nanocage with CIP. Quantum theory of atoms in molecules (QTAIM) analysis of the complexes of Zn12O12 nanocage with CIP. Natural bond orbital analysis of Zn12O12 nanocage and its complexes with CIP Natural bond orbital (NBO) analysis was used to determine charge transfer that took place between the CIP drug and the Zn12O12 nanocage. The NBO analysis was performed at the B3LYP/ 6-31G(d, p) level on the optimized Zn12O12-CIP complexes using the NBO code implemented in the Gaussian 16 software (https://seagrid.org/). The values of the charge transfer between the CIP molecule and the Zn12O12 nanocage are given in Table 1. Electronegative atoms (oxygen and fluorine in the CIP drug and Zn12O12 nanocage) have partial negative charges in the isolated molecules (monomers), and the hydrogen atoms in CIP and the zinc atoms in the nanocage have partial positive charges. On complexation, the partial charges on the oxygen atom of CIP moieties increase, whereas the charges on zinc atoms decrease, suggesting charge redistribution from the nanocage toward the CIP molecule. The charge of the oxygen atom in CMP-1 increases to -0.65 e from − 0.56 e upon complexation, witnessing a charge transfer of -0.09 e. Similarly, the oxygen atoms of the carbonyl groups of the CIP molecule in CMP-2 and CMP-3 show an increase in charge from − 0.56 e and − 0.57 e to -0.66 e and − 0.71 e, respectively. On the other hand, the fluorine atom in CMP-4 shows a small change in charge, suggesting weak interaction and lower stability of CMP-4 among the optimized complexes. Conversely, the charge on the oxygen atom of the hydroxyl (O-H) group of the CIP molecule in CMP-5 changes to -0.81 e from − 0.75 e, indicating significant charge redistribution of charge during adsorption. Overall, the charge transfer analysis reveals the existence of interactions between CIP and Zn12O12 nanocage. The findings of the NBO analysis suggest that Zn12O12 nanocage is a suitable candidate for the adsorption of CIP. Moreover, the obtained results of NBO are in good agreement with the calculated Eads, witnessing the sensing capability of the nanocage for the drug. Electron density difference (EDD) analysis To further justify the stability of the interactions between the Zn12O12 nanocage and the CIP molecule, EDD analysis was employed to support the results of natural bond orbital (NBO) analysis. The EDD isosurfaces shown in Fig. 8 depict the accumulation and depletion of electron density and provide insight into the nature of bonding between the adsorbent and the adsorbate. The EDD plots of all five complexes show pronounced blue and green isosurfaces regions at the interacting sites between the Zn12O12 nanocage and CIP molecule, indicating the presence of interactions between the interacting species. The presence of large green loops in CMP-1, CMP-2, and CMP-3 covers the interacting atoms, which indicates electron redistribution and maximum orbital overlap. The maximum orbital overlap agrees with high Eads values, supporting the charge transfer analysis. These findings suggest the stability of the complexes and demonstrate the capability of Zn12O12 nanocages for the removal of the CIP molecule from wastewater. It should be noted that EDD and NBO analyses provide theoretical insight into charge redistribution and intermolecular interactions but do not directly establish adsorption efficiency under realistic environmental conditions. Fig. 8Electron density difference (EDD) analysis of the complexes of Zn12O12 nanocage with CIP. Electron density difference (EDD) analysis of the complexes of Zn12O12 nanocage with CIP. Effect of solvent and dipole moment analysis The effects of solvent environment on the interaction behavior of the Zn12O12 nanocage and the CIP molecule were analyzed using the polarizable continuum model (PCM) in which the solvent is represented as a continuous polarizable dielectric medium without explicitly calculating individual solvent molecules. The PCM model was applied to check the influence of water as a solvent. The adsorption energy values in the gas phase and the aqueous medium are tabulated in Table 4. A slight variation in adsorption energies is observed in the aqueous phase compared to gas-phase calculations. In most cases, a decrease in adsorption energy is observed, indicating partial solvent screening of the interaction strength. However, for CMP-1 and CMP-4, a slight increase is observed, suggesting that solvent polarity may stabilize specific interaction geometries within the PCM approximation. CMP-3 showed the most favorable adsorption energy of − 43.20 kcal/mol, indicating that CMP-3 remains the most strongly interacting configuration in the aqueous phase within the PCM model. Further justification for the existence of interactions between the adsorbent and adsorbate was obtained by employing dipole moments analysis. The isolated adsorbent and the CIP molecule have dipole moments of about 0.000173 and 9.70 Debye, respectively. The dipole moments were increased significantly for all configurations, indicating the existence of interactions between the nanocage and drug, as given in Table 4. This increase in polarity implies the increased charge redistribution, suggesting the adsorption capability of the adsorbent for the elimination of the CIP drug from wastewater. Table 4Adsorption energy (Eads) values in gas phase and solvent, thermodynamics, and dipole moment (DM) analyses.SystemEads (gas medium)Eads (water medium)DM (Debye)∆H, Kcal/mol∆G, Kcal/molZn12O12––0.000173––Ciprofloxacin––9.70––CMP-1− 39.37− 41.1920.43− 37.04− 19.44CMP-2− 39.32− 38.0216.29− 37.89− 17.89CMP-3− 48.99− 43.208.58− 44.77− 27.45CMP-4− 23.60− 23.9411.71− 20.71− 9.88CMP-5− 33.48− 27.498.74− 31.12− 13.52 Table 4Adsorption energy (Eads) values in gas phase and solvent, thermodynamics, and dipole moment (DM) analyses.SystemEads (gas medium)Eads (water medium)DM (Debye)∆H, Kcal/mol∆G, Kcal/molZn12O12––0.000173––Ciprofloxacin––9.70––CMP-1− 39.37− 41.1920.43− 37.04− 19.44CMP-2− 39.32− 38.0216.29− 37.89− 17.89CMP-3− 48.99− 43.208.58− 44.77− 27.45CMP-4− 23.60− 23.9411.71− 20.71− 9.88CMP-5− 33.48− 27.498.74− 31.12− 13.52 Adsorption energy (Eads) values in gas phase and solvent, thermodynamics, and dipole moment (DM) analyses. Thermodynamics analysis The thermodynamic analysis indicates that the calculated values of enthalpy change (ΔH) and Gibbs free energy change (ΔG) for all Zn12O12–CIP complexes are negative (Table 4), suggesting that the adsorption process is exothermic and thermodynamically favorable under the studied conditions. The negative ΔG values further indicate the spontaneity of the interaction. Similarly, no imaginary frequencies were observed for the complexes, reflecting the stability of the complexes. These results are consistent with the energetic and electronic structure analyses, supporting the presence of favorable intermolecular interactions between the adsorbent and adsorbate. Recovery time calculation and molecular dynamics study The regeneration capability of the Zn12O12 nanocage is an important factor for evaluating adsorption reversibility. The recovery times (τ) of all five complexes were calculated at 298.15 K and 350 K. At 298.15 K and depicted in Fig. 9, the recovery times were found to be in the range of 2.00 × 105 to 8.20 × 1023 s, indicating very strong adsorption interactions, particularly for CMP-3, which exhibits extremely slow desorption kinetics. Upon increasing the temperature to 350 K, a decrease in recovery time is observed for most complexes, suggesting that thermal energy facilitates partial desorption of CIP from the nanocage surface. However, even at elevated temperature, several complexes still show very long recovery times, indicating that spontaneous regeneration under ambient conditions is unlikely for strongly bound configurations. Fig. 9Recovery time graph of the complexes of Zn12O12 nanocage with CIP. Recovery time graph of the complexes of Zn12O12 nanocage with CIP. To further evaluate thermal stability, AIMD simulations were performed for CMP-4 at 500 K, which showed no structural collapse within the simulation time scale as shown in Fig. 10, suggesting that the nanocage framework remains structurally stable under elevated temperature conditions. Fig. 10AIMD analysis of CMP-4. AIMD analysis of CMP-4. Overall, these results indicate strong adsorption affinity of Zn12O12 toward CIP; however, the large recovery times suggest that desorption and regeneration may require external stimuli such as heating or other regeneration strategies. Therefore, the reusability of the system should be interpreted cautiously, and experimental validation is required to assess practical applicability in wastewater treatment. In this work, a comprehensive density functional theory (DFT) investigation was carried out to explore the interaction behavior of ciprofloxacin (CIP) with a Zn12O12 nanocage. Various analyses, including geometry optimization, adsorption energy calculations, frontier molecular orbital (FMO) analysis, global reactivity descriptors, natural bond orbital (NBO) analysis, reduced density gradient (RDG), quantum theory of atoms in molecules (QTAIM) studies, electron density difference (EDD) analysis, and thermodynamic evaluations, were employed to gain detailed insight into the adsorption mechanism. The calculated adsorption energies indicate that CIP interacts strongly with the Zn12O12 nanocage, with all investigated configurations exhibiting negative adsorption energies, suggesting thermodynamically favorable adsorption. Among all studied systems, CMP-3 shows comparatively stronger interaction, as supported by adsorption energy, charge transfer, and electronic structure analyses. The FMO and DOS results further indicate noticeable modulation of the electronic properties of the nanocage upon adsorption, reflecting significant orbital interaction between CIP and Zn12O12. Global reactivity descriptors and NBO analysis suggest charge redistribution between CIP and the nanocage, indicating electron transfer–assisted interaction. RDG, NCI, and QTAIM analyses collectively support the presence of mixed interaction characteristics, including electrostatic, hydrogen bonding, and partial covalent contributions. Thermodynamic parameters further confirm that the adsorption process is spontaneous and exothermic under the studied theoretical conditions. Recovery time analysis indicates that strong adsorption leads to significantly long desorption times for several complexes, particularly at room temperature, while elevated temperature partially reduces recovery times. However, the large magnitude of recovery times for some configurations suggests that spontaneous desorption may be limited, and regeneration could require external energy input or additional treatment strategies. Although the present study suggests favorable interaction of CIP with Zn12O12 nanocage at the theoretical level, it is important to emphasize that all findings are based on computational models. The PCM solvent model, adsorption simulations, and electronic structure analyses provide valuable molecular-level insight; however, they do not fully capture real environmental conditions such as pH effects, competing ions, aggregation of the adsorbent, explicit solvent interactions, or material stability in complex wastewater systems. Therefore, while Zn12O12 nanocage demonstrates promising theoretical affinity toward ciprofloxacin, its practical applicability for wastewater treatment remains to be experimentally verified. Future studies involving experimental synthesis, adsorption measurements, and regeneration tests are necessary to validate its real-world performance. Computational details Density functional theory (DFT)43 is accurate, reliable, and has a high level of computational efficiency when it comes to the study of the overall structure of materials and their electronic characteristics44–48. The optimization of the geometries of CIP as well as Zn12O12 nanocage and the complexes was performed using the GAMESS49 software (https://chemcompute.org/) with the use of the B3LYP theory and 6-31G(d, p) basis set. All systems were treated in the natural state with zero charge and singlet multiplicity. To accurately account for long-range dispersion interactions, Grimme’s50 DFT-D3 dispersion correction method was employed throughout the optimization of the adsorbate, adsorbent, and complexes using the GAMESS49 software (https://chemcompute.org/). The adsorption energy values of the complexes in the gas phase and aqueous conditions were obtained using the following equation. 1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{ads}={E}_{complex}-({E}_{M1}+{E}_{M2})$$\end{document} Ecomplex is the overall energy of the CIP-Zn12O12 system, and EM1 and EM2 are the energies of the free CIP molecule and Zn12O12 nanocage, respectively. The global reactivity descriptors51, namely, ionization potential (IP), electron affinity (EA), electronegativity (χ), global hardness (η), global softness (S), chemical potential (µ), and electrophilicity index (ω), were computed for the CIP and Zn12O12 molecules as well as their corresponding complexes using the equation. 2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:IP\approx\:\:-{E}_{HOMO}$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:EA\approx\:-{E}_{LUMO}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upchi\:}\:=\:-{\upmu\:}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upeta\:}=\frac{{E}_{LUMO}-{E}_{HOMO}}{2}$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:s=\frac{1}{2{\upeta\:}}$$\end{document}7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upmu\:}\:=\:\frac{{E}_{LUMO}+{E}_{HOMO}\:}{2}$$\end{document}8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:=\frac{\mu\:}{2{\upeta\:}\:}}^{2}$$\end{document} Frontier molecular orbital (FMO)48 analysis was carried out employing Gaussian 16 (https://seagrid.org/), and the HOMO and LUMO energies and the corresponding HOMO-LUMO plots were obtained using the checkpoint (.chk) file with the help of Gauss View 6.0 software (https://gaussian.com/gaussview6/). Density of states (DOS) analysis was carried out with GaussSum 3.0 using the output files of the GAMESS software to investigate the electronic properties and sensing mechanism. The following expression was used to calculate the HOMO-LUMO energy gap (Eg).9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{E}_{g}={E}_{LUMO}-{E}_{HOMO}$$\end{document} ELUMO and EHOMO are the energies of the lowest and highest unoccupied and occupied molecular orbitals, respectively. The wavefunction (.wfn) files were generated using the Gaussian 16 (https://seagrid.org/) and then utilized for NCI, RDG, QTAIM, and EDD analyses. Visualization for non-covalent interactions (NCI) was carried out in Visual Molecular Dynamics (VMD) 1.9.452 software (https://www.ks.uiuc.edu/Research/vmd/), whereas the files (.cube) required for NCI analysis were obtained from the wavefunction (.wfn) files using the Multiwfn 3.653 software (http://sobereva.com/multiwfn/). Reduced density gradient (RDG) analysis was performed using the wavefunction (.wfn) files, and the graphs were obtained with the help of Multiwfn 3.6 (http://sobereva.com/multiwfn/). The topological parameters of electron density were also studied based on the quantum theory of atoms in molecules (QTAIM) method to determine the bond critical point (BCP)54 by utilizing Multiwfn 3.6 software (http://sobereva.com/multiwfn/) using the wavefunction (.wfn) files. The charge transfer (NBO) analysis was performed55 using the NBO code as implemented in the Gaussian 16 (https://seagrid.org/), and the results were obtained employing Gauss View 6.0 software (https://gaussian.com/gaussview6/). Electron density difference (EDD)56 analysis was performed from the wavefunction (.wfn) files using Multiwfn 3.6 software (http://sobereva.com/multiwfn/), and the plots were generated employing Gauss View 6.0 software (https://gaussian.com/gaussview6/) to visualize the redistribution of charge during interaction. The Hessian calculations were performed at the same DFT level of theory and basis set implemented in GAMESS software (https://chemcompute.org/) to find the stability of the complexes. The thermodynamics parameters57 such as enthalpy changes (∆H)58 and Gibbs free energy (∆G)59, were calculated employing GAMESS software (https://chemcompute.org/) using the following relations.10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{H}_{ads}={H}_{Complex}-{H}_{(M1+M2)}$$\end{document}11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{G}_{ads}=\varDelta\:{H}_{ads}-T\varDelta\:{S}_{ads}$$\end{document}12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{G}_{ads}=\varDelta\:{H}_{ads}-T\left[\right({S}_{\left(complex\right)}-({S}_{adsorbent}+{S}_{adsorbate})]$$\end{document} where ∆H and ∆G are the total electronic and thermal enthalpy and Gibbs free energy at standard conditions (298.15 K and 1 atm), and S is the entropy. The dipole moment (DM) analysis was performed using Gaussian 16 software (https://seagrid.org/). The influence of solvent on the adsorption of CIP was simulated using the Polarizable Continuum Model (PCM)60 with water as solvent, utilizing the GAMESS software (https://chemcompute.org/). The assessment of adsorbent regeneration (recovery time)61 was performed at 298.15 K and 350 K to determine the recyclability as well as thermal stability of the adsorbent employing the equation. 13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\uptau\:}=v}_{0}^{-1}Exp(-{E}_{ads}/kT)$$\end{document} Here, V0−1 denotes the attempt frequency, Eads signifies the adsorption energy, k stands for Boltzmann’s constant, and T represents the temperature. The stability of the complexes was checked using the least stable complex (CMP-4) as a reference, employing ab-initio molecular dynamics (AIMD)62 with the help of the Dmol3 code using the PBE functional along with the generalized gradient approximation (GGA). The basis set employed is double numerical plus polarization (DNP), and the graph was drawn using Microsoft Excel 2019. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1