Facile fabrication of novel magnetic chitosan@Ag-MWCN nanocomposite for the adsorptive removal of ciprofloxacin from aqueous solutions

Characterization of MC@Ag-MWCN

FTIR analysis

The FT-IR research of the chitosan nanoparticles validates the presence of the bands at −1898.52 cm⁻¹ illustrates the existence of CH bond in the monosaccharide ring, 1078.01 cm⁻¹ for C–O bond, 1157.28 cm⁻¹ for C–O–C asymmetric stretching bond , 1320.73 cm⁻¹ for C–N bond in type three amide and 1593.19, 1654.29 cm⁻¹ indicate the stretching vibration of the C = O bond in the amide group, while a peak at about 3434.23 cm⁻¹ can be related to the stretching vibration of the N–H and O–H bands (Fig. 2) ²⁹. The FT-IR analysis of the functionalized MWCNTs reveals the existence of defect sites on their surface. Bands observed at 1637 cm⁻¹ indicate > C = C bonds, around 1008 cm⁻¹ for > C = O groups, 2854 and 2930 cm⁻¹ for = CH₂ groups, and 3439 cm⁻¹ corresponding to OH^– functional groups. Additionally, a peak at approximately 1400 cm⁻¹ may be attributed to the carboxylic group -COO attached to the MWCNTs (see Fig. 2)²⁶. The FT-IR spectra of the magnetic nanocomposite validate the existence of > Fe = O at 448.46 cm⁻¹, > C = O at 636.26 and 887.34 cm⁻¹, C–N at 1401.22 cm^–1 and the peak at 3387.73 can be considered related to the stretching vibration of the hydroxyl group (OH) or the NH band of the magnetic nanocomposite (Fig. 2)²⁹. The result of final FTIR spectra of MC@Ag-MWCN result show the peak 485.81 cm⁻¹, 634.82 and 886.27 cm⁻¹, 1079.06 cm⁻¹, 1385.61 cm⁻¹ which can be associated to Fe–O , C–O for (CNT), C–O for (Chitosan) and CH3 bond (Chitosan) respectively. Also, this FTIR spectra confirmed presence C = O for Chitosan, AgNps and CNT at 1627.70 cm⁻¹, C-H for AgNps at 2378.10 and 2920.75 cm⁻¹ N–H for Chitosan, AgNps and CNT at 3433.02 cm⁻¹. (Fig. 2). Therefore, it can be concluded that the synthesis of MC@Ag-MWCN nanocomposite has been done successfully.

FTIR spectra of Chitosan, functionalized MWCNTs, magnetic Fe₃O₄ and MC@Ag-MWCN nanocomposite.

XRD analysis

The MC@Ag-MWCN crystal structure was investigated using XRD analysis. Figure 3 shows the diffraction pattern of functionalized carbon nanotubes, Ag and chitosan nanoparticles. The diffraction pattern of MWCNTs shows three peaks at 26.2°, 44.7°, and 5.46°, which are related to carbon nanotubes. The diffraction pattern of Ag shows four peaks at 38.1°, 64.4°, 44.5°, and 78.4°. Also, this Fig. 3. confirms the presence of the bands the index peak observed at 11.6° is related to chitosan particles ^{26, 29}

The XRD pattern of MC@Ag-MWCN used for CIP removal.

FE-SEM and TEM analysis

To investigate the morphology of hybrid nanostructures, FESEM microscope was used. The typical images of FE-SEM which are presented in Fig. 4. These images (4a, 3b) confirm the filament components of carbon nanotubes and particle structures of chitosan and Ag nanoparticles. Figure 4a shows the image of CNT/Ag/Cs triple hybrid, in which Ag nanoparticles are almost uniformly seated on the body of nanotubes. As can be seen in the figure, there is no accumulation and clumping of particles. In addition, the Nano meter of Ag particles and the tubular structure of carbon nanotubes could be discerned. Also, in Fig. 4b, the relatively uniform accumulation of spherical chitosan particles is completely visible. To further investigate the morphology of CNT/AG/CS nanocomposite, it was examined by TEM microscope. Figure 4c shows the CNT/AG/CS ternary hybrid, confirming the deposition of Ag nanoparticles on the surface of carbon nanotube filament. This image confirms the presence of particles on the body of carbon nanotubes²⁶. It seems that the darker particles correspond to silver nanoparticles and the lighter particles represent chitosan particles^{29, 30}. Darker particles (Ag) in Fig. 4d can be seen as semi-spherical and curved, and lighter particles (CS) are completely spherical without angles²⁶.

Field emission scanning electron microscopy (a, b) and TEM images of chitosan /Ag @Magnetic Fe₃O₄MWCN nanocomposite (c, d).

Parametric study

CIP concentration

Initial concentration of the pollutant is a key factor affecting the adsorption process. This research investigated initial concentrations of CIP ranging from 25 to 100 mg/L. As illustrated in Fig. 5a, the initial concentration of CIP negatively impacts removal effectiveness. Specifically, as the CIP concentration rises from 25 to 100 mg/L, the removal effectiveness reduces from 92 to ~ 72.7%. This reduction in removal efficiency with higher initial antibiotic concentrations can be explained by the fact that the number of active adsorption sites remains constant with a constant amount of adsorbent. Consequently, as the concentration of the pollutant rises, the number of molecules in the reaction medium increases, leading to a decrease in removal efficiency due to the saturation of available adsorption sites³¹. Mohammadi Amini et al. and Alidadi et al., conducted a study that applied magnetic-chitosan nanocomposite and Chitosan/Zeolite composite in the removal of antibiotics. Their result indicated that the removal efficiency decreases with the increasing concentration of CIP and metronidazole^{32, 33}.

Adsorption of CIP by MC@Ag-MWCN as a function of (a) CIP concentration (b) pH, (c) adsorbent dose, and (d) contact time.

pH

The pH of the solution is a crucial factor in purification processes such as adsorption. The surface charge of the adsorbent, the ionization of different pollutants, the distribution of functional groups on the adsorbent’s active sites, and the structure of the antibiotic molecule can be affected by pH. Essentially, pH impacts the overall chemistry of the solution and the interactions at the adsorbent surface³¹. In this study, CIP removal examined as a function of pH between 5 and 11. Acording to the results the removal approach by MC@Ag-MWCN shows the best effectiveness at pH 7–9. The highest removal observed when pH reaches about pH 9. (Fig. 5b). CIP molecule are predominantly cationic at pH < 5 by the protonation of its amine groups. Conversely, at pH > 9 CIP shifts to an anionic form as the carboxylic group loses protons. Therefore, in this pH range between 7 and 9, most of the CIP molecule in the aqueous solution is uncharged. This finding agrees with the CIP removal by other adsorbents in pH between 6 and 8.5^{32, 34, 35}

The impacts of adsorbent dose of Magnetic chitosan /Ag @ MWCN nanocomposite

Examining the impact of adsorbent mass on adsorption processes is a critical aspect that needs careful consideration. The adsorbent dosage varied between 0.2 and 0.8 g/L, and its impact on the removal efficiency of ciprofloxacin was assessed. According to the obtained data at a fixed concentration of antibiotics, the removal effectiveness also increased with the rise of the adsorbent dose. As Fig. 5c shows, when the antibiotic concentration remains constant at 25 mg/L and the adsorbent adjusted to 0.2 g/L, the removal performance reached 76.3%, and ~ 92.5% at a dose of 0.6 g/L. The reason for this can be stated that in a fixed antibiotic concentration, with the increase of the adsorbent dose, the proportion of the surface’s active sites on adsorbent to the molecules of the substance increases. A low quantity of adsorbent causes a reduction in the ratio of active sites to adsorbate molecules, leading to decreased adsorption efficiency. Conversely, when the adsorbent amount exceeds the optimal level, the adsorption capacity declines from its peak value of 25.52 mg/g. This decrease occurs because, at higher adsorbent doses, the available active sites become fully utilized. As a result, any additional adsorbent does not enhance the adsorption capacity. This observation can be attributed to the saturation of the adsorbent’s surface, where excess adsorbent does not contribute further to the adsorption process. Mohammadi Amini et al. (2019) also showed an increasing the magnetic-chitosan nanocomposite adsorbent from 0.1 to 0.5 g/L, the removal efficiency for tetracycline increased from 72.24 to 90.22% at 90 min³².

Contact time

An essential factor in the application of an adsorption system is providing adequate contact time to ensure optimal performance. In this study, we examined contact time ranging from 20 to 80 min and analyzed its impact on the process. Figure 5d shows that CIP removal increased from 74.23 to 92.1% once mixing time increased from 20 to 60 min. With the increase of time from 60 to 90 min, the efficiency decreased again and reached 82%.

The presumed adsorption mechanism

Isotherms study

Adsorption isotherms describe mathematical relationship between experimental amount of substance in gas or liquid phases and adsorbent surface at constant temperature. Herein, the experimental equlibrium data fit with common isotherms such as Langmuir, Freundlich, Dubinin–Radushkevich and Temkin equations.

Freundlich isotherm model

In this experimental model, the relationship between the residual concentration of the adsorbent material and the adsorption capacity is defined as follows³⁶.

Freundlich isotherm illustrates a relationship between non-ideal and reversible adsorption. This experimental model can be used for multilayer adsorption with heterogeneous heat distribution and adsorption on heterogeneous surface. By taking the logarithm from both sides of the relationship, it can be converted into a linear form and by plotting (ln(qe) in terms of (ln(Ce), the parameters kF and n according to the slope and width from the origin of the straight line with the equation given in the following relationship Gained³⁷.

Langmuir isotherm model

In the Langmuir isotherm model, adsorption takes place exclusively at a finite number of distinct and uniform sites. In each of these active sites, there is no subsequent reaction and no space barrier between the adsorbent and the adsorbate. Therefore, the Langmuir isotherm assumes that adsorption occurs uniformly across the adsorbent’s surface, with adsorbent molecules having constant enthalpy and activation energy. This model also assumes that once a molecule is adsorbed, it remains fixed on the surface without displacing other adsorbed molecules. In this framework, the focus is on the portion of the surface that becomes saturated and is no longer capable of adsorbing additional molecules at equilibrium. In this model, adsorption is assumed to be a single layer and expressed as the below relationship³⁸.

where Ce is the equilibrium concentration of the solute in the solution in the context of (qe), (mg/L) and qmax respectively, the equilibrium capacity of adsorption and the maximum capacity of adsorption in (mg/g). The constant value of kL in these relationships is the tendency of the soluble component towards the adsorbent in terms of (L/mg) and the distribution constant is in terms of (L/g). At low concentrations, this model turns into a linear model and acts like Art’s law.

Temkin isotherm model

The Temkin isotherm equation is becoming more commonly used to analyze and correlate adsorption isotherms for water pollutants. This model includes a factor that clearly explains the reason for the interaction between the adsorbent and the adsorbed substance³⁹. Temkin isotherm is presented in Eqs. (6)–(7)

\({b}_{T}\)= Temkin isotherm constant (kJ mol⁻¹).

\({A}_{T}\)= Binding constant at equilibrium (L g⁻¹)

Dubinin–Radushkevich isotherm model

Dubinin–Radushkevich isotherm model is an empirical adsorption model specifically suited for intermediate concentrations of the adsorbate. It is commonly employed to distinguish between physical and chemical adsorption of metal ions. The Dubinin–Radushkevich isotherm is expressed by the Eqs. (8)–(9)^{40, 41}.

From the linear form fits of the applied four isotherm models (Fig. 6a–d) and the provided parameters (Table 1), it could be inferred that the values of R² for Langmuir, Friedlich, Temkin and Dobin-Radeshkovij isotherms were obtained 0.887, 0.73, 0.884 and 0.777, respectively. The results showed that the adsorption of CIP onto MC@Ag-MWCN is more consistent by the Langmuir isotherm model (R² = 0.887). Therefore, the q_max value calculated from the Langmuir adsorption isotherm model was utilized to compare the different adsorbents that were applied for CIP removal. Based on the literature review provided in Table 2, the value of q_max = 31.26 mg/g is relatively high and categorizes the adsorbent MC@Ag-MWCN as effective in the process of CIP adsorption.

The isotherms study of the CIP adsorption by MC@Ag-MWCN: (a) Langmuir, (b) Freundlich, (c) Dubinin–Radushkevich and (d) Temkin.

Kinetic study

Synthetic models are utilized to examine the adsorption mechanism and control the reaction speed steps such as mass transfer and the progress of chemical reactions⁴⁷. This research examines the kinetic model of ciprofloxacin adsorption by the MC@Ag-MWCN, and various models such as pseudo first order, pseudo second order and interparticle diffusion were used. The coefficients of the pseudo-first-order kinetic equation and its correlation coefficient were measured from the plot of ln (qeq-qt) versus t. Also, the linear graph of t/qeq versus t was drawn to determine the coefficients of the pseudo-second-order kinetic equation and determine the corresponding correlation coefficient. The attained parameters from the kinetics plots (Fig. 7a–c), reflected the chemisorption interactions between CIP and MC@Ag-MWCN, where the experimental qe are near to the computed qe from PSO at the varied CIP concentrations(Table 3) (Fig. 7b). Furthermore, the correlation coefficients of the PSO curves are more than those of PFO. According to the kinetic results of ciprofloxacin adsorption, the correlation coefficient for PFO, PSO, and interparticle diffusion was obtained as 0.531, 0.994, and 0.753, respectively. Therefore, it can be inferred that the adsorption of ciprofloxacin follows PSO kinetic model (Table 3) (Fig. 7b). This shows that in this study, chemical adsorption is the rate-limiting step and controls the kinetic of adsorption. Also, this compliance shows that the nanocomposite has a smooth and regular surface with uniform bonding energies. This finding is consistent with the result obtained in the earlier studies on chitosan-biochar hydrogel adsorbent, modified chitosan-biochar hydrogel adsorbent and magnetite–imprinted chitosan nanocomposite^48,49,50.

Pseudo-first-order (a), pseudo-second-order (b) and intraparticle diffusion models (c) fitted to CIP uptake by MC@Ag-MWCN.

Adsorption mechanism of MC@Ag-MWCN nanocomposite

Figure 8 presents a schematic diagram that illustrates how the MC@Ag-MWCN nanocomposite removes ciprofloxacin (CIP). The removal of ciprofloxacin (CIP) using the MC@Ag-MWCN nanocomposite occurs through a combination of physical and chemical interactions, facilitated by the composite’s structural and chemical properties. At a pH range of 7–9, CIP molecules predominantly exist in a neutral form, enhancing their affinity to the adsorbent surface through electrostatic interactions. At acidic pH (< 7), CIP becomes positively charged, while the surface groups of MC@Ag-MWCN may also be protonated, potentially reducing electrostatic attraction. At alkaline pH (> 9), the deprotonation of functional groups on both CIP and the adsorbent leads to repulsion, reducing removal efficiency. Functional groups such as –NH2, –OH, and –COOH on chitosan and carbon nanotubes act as hydrogen bond donors/acceptors. These groups form hydrogen bonds with the hydroxyl and amine groups of CIP, enhancing adsorption³¹. The aromatic structures of multi-walled carbon nanotubes (MWCNTs) and the aromatic rings in the CIP molecule facilitate π–π stacking interactions, stabilizing the adsorption process. Silver nanoparticles (AgNPs) in the nanocomposite improve the adsorptive capacity by forming strong interactions with CIP, including coordination bonds. The pseudo-second-order kinetic model, which best fits the experimental data, suggests that chemical adsorption plays a significant role. The hydrophobic surfaces of MWCNTs interact with the non-polar regions of the CIP molecule, aiding in adsorption. Oxidized MWCNTs contain carboxylic, hydroxyl, and carbonyl groups, which provide active sites for binding CIP through coordination and hydrogen bonding. Chitosan contributes through amine and hydroxyl groups, further enhancing the adsorption mechanism^{26, 48}. The removal process of CIP by MC@Ag-MWCN aligns with findings in adsorption studies involving chitosan-based and carbon nanotube-functionalized nanocomposites. The Langmuir isotherm model with a maximum adsorption capacity of 31.26 mg/g indicates monolayer adsorption on homogeneous sites. This is consistent with studies that employed chitosan/graphene composites for antibiotic removal, where similar adsorption capacities and mechanisms were observed⁵¹. The pseudo-second-order kinetic model with a high correlation coefficient (R² = 0.994) indicates chemisorption dominates the adsorption mechanism. Comparable results have been reported for magnetic chitosan/zeolite composites used for antibiotic removal⁵². The functionalization of MWCNTs enhances adsorption efficiency through defect sites and functional groups that facilitate interactions with CIP. Ag nanoparticles improve removal efficiency through enhanced electronic interactions and antibacterial properties²⁹.

Adsorption mechanism of MC@Ag-MWCN nanocomposite on the CIP removal.