Characteristics of Fe–S nanoparticles and Fe–S/CuS nanocompositeXRD analysisAs shown in the XRD patterns of Fe–S nanoparticles Fig. 1, characteristic peaks of α-Fe0 were observed at 2θ = 35.70°, 43.30°, 57.21°, 62.80°, and 74.28°, with the peak at 35.70° suggesting the formation of magnetite37. The distinct peak at 44° signifies the presence of Fe. Furthermore, weak iron oxide peaks were detected at 2θ = 18.64°, 30.34°, and 53.69°, suggesting that the surfaces of the nanoparticles underwent partial oxidation during the synthesis. peaks in Fig. 1 were observed at 29.55°, 30.33°, 32.94°, 48.22°, 53.04°, and 59.58° (2θ), correlating with the formation of CuS18. This further indicates the successful loading of CuS onto Fe–S nanoparticle25.Figure 1The XRD patterns of Fe–S nanoparticles and Fe–S/CuS nanocomposite.FESEM analysisAs shown in the SEM image Fig. 2, most of the particles are nano-sized. According to the histogram chart, Fe–S particle sizes range from 49 to 118 nm with an average diameter of 83.5 nm. Additionally, Fe–S/CuS nanocomposite particle sizes range from 75 to 96 nm with an average diameter of 85.5 nm. Images showed that Fe–S and Fe–S/CuS nanocomposite are similar in appearance and size. The particles are distributed in a regular and uniform spherical shape, which is important for improving their performance. The dark points in Fig. 2 show the iron particles well and uniformly, with CuS seated and stabilized on the substrate. Due to Fe–S and Fe–S/CuS, there is a possibility of accumulation within them25.Figure 2FESEM Analysis of Fe–S nanoparticles and Fe–S/CuS nanocomposite.FTIR analysisThe FTIR spectra were analyzed to identify the functional groups present within the 500–4000 cm−1 range are shown in Fig. 3. The absorption peaks observed at 3413.38 and 3317.03 correspond to O–H groups bonded to the ring carbon38. Furthermore, the absorption peak at 2921.86 cm−1 in the Fe–S/CuS nanocomposite spectrum was attributed to the vibration of the Fe-Cu bond, which overlapped with the peaks corresponding to O–H functional groups38. The distinct absorption peaks of Fe–S/CuS nanocomposite at 1623 cm−1 and 1123.74 cm−1 corresponded to the stretching vibration of carbonyl groups (S = S, C = O), and COO − , respectively18,27,39. The peak at 588.69 cm−1 is likely characteristic of the vibration associated with the Fe–S bond, indicating the presence of Fe–S nanoparticles40. A peak at approximately below 500 cm−1 may indicate the presence of the iron-CuS composite, attributed to the stretching mode of the Cu–S bond with sulfur as the heteroatom40,41.Figure 3FTIR spectra of Fe–S nanoparticles and Fe–S/CuS nanocomposite.VSM analysisAs shown in Fig. 4, the magnetic saturation values for each of these compounds were 33.16 emu g−1, and 3.49 emu g−1, respectively, indicating their suitability for magnetic separation42. Based on the VSM analysis results in Fig. 4, it can be concluded that the Fe–S nanoparticles exhibited the highest magnetic potential41. Loading CuS onto the Fe–S nanoparticles partially decreased their magnetization significantly, diminishing their magnetic strength41.Figure 4VSM analysis of Fe–S nanoparticles and Fe–S/CuS nanocomposite.TEM analysisAs shown in Fig. 5 the particle size of the magnetic nano composite, created using Fe–S nanoparticles, falls within the nano scale range, with spherical particles. The irregular cumulative state of the particles can also be attributed to the high magnetic properties of this nanocomposite18. At higher magnification in the TEM image in Fig. 5, it becomes evident that numerous interconnected Fe–S/CuS nanocomposites exhibit core–shell structures and agglomerate into chain-like formations due to strong surface energies38.Figure 5TEM Analysis of Fe–S nanoparticles and Fe–S/CuS nanocomposite.BET analysisThe specific surface area analysis was conducted using the Brunauer–Emmett–Teller (BET) method, with nitrogen adsorption on an ASAP2020 surface instrument. The nitrogen adsorption/desorption isotherms of Fe–S/CuS nanocomposites exhibit a unique type-IV isotherm at 77 K. This isotherm suggests that the nanocomposite has a predominantly mesoporous distribution in its specific surface area and porous structure18. The BET specific surface area results for Fe–S/CuS nanocomposites indicate 37.86 m2g−1, exceeding the surface area of Fe–S particles, which is only 12.70 m2g−1. This observation is consistent with the findings reported by Li et al.18,37. This property makes the Fe–S nanocomposite a preferred choice for use as an adsorbent. The pore size distribution of BJH, derived from the adsorption branch for both Fe–S and Fe–S/CuS nanocomposite, is shown in Fig. 6 and Table 2. The average pore size of Fe–S/CuS nanocomposite is approximately 12.96 nm, suggesting the presence of fairly uniform pores in the Fe–S/CuS nanocomposite material37. It was also found that the Fe–S sample had an average pore size of 15.22 nm, which followed a meso-Prussian pore size distribution18. The pore volume and pore size of Fe–S nanoparticles and Fe–S/CuS nanocomposite, calculated using the BET isotherm adsorption curve equation, are 0.04 and 0.12 cm3g−1 respectively, the findings suggest that the pores in Fe–S result from the aggregation of particles37. Furthermore, CuS-coated Fe–S nanoparticles can increase surface defects in the sample, thereby providing abundant adsorption sites for TBBPA molecules on the specific surface of the Fe–S/CuS nanocomposite.Figure 6Adsorption/desorption isothermand and BJH-Plot of Fe–S nanoparticles and Fe–S/CuS nanocomposit.Table 2 BET surface area and pore volume parameters of Fe–S nanoparticles and Fe–S/CuS nanocomposite.EDS map analysisTo further support the SEM, additional evidence has been obtained through EDS analysis conducted on the prepared samples. The Fig. 7 depicts the EDS spectra of Fe–S/CuS nanocomposite and CuS samples doped with Fe. The nanoparticles consist of iron with copper present on the surface. This suggests the co-existence of Fe and Cu ions in the Fe–S/CuS nanocomposite. The EDS in Fig. 7 of the Fe–S/CuS nanocomposite confirms the homogeneous distribution of the elements Fe, Cu, and S in the products. This indicates the uniform dispersion of FeS and CuS within the Fe–S/CuS nanocomposite43.Figure 7EDS spectra Fe–S nanoparticles and Fe–S/CuS nanocomposite.Response surface methodology (RSM)The results obtained from experimental data and model predictions regarding TBBPA removal through the adsorption process on Fe–S/CuS nanocomposite were analyzed using a Central Composite Design. A total of 54 runs were conducted according to Response Surface Methodology and the outcomes are presented in Table S1 in the supplementary file. The final model, represented in Eq. (3), forecasts the removal percentage values (R %) of TBBPA:$$\begin{aligned} {\text{Removal}}\;{\text{TBBPA}}\;\left( \% \right) \, = & + 50.78{-}8.69X_{1} + 0.8793X_{2} {-}6.51X_{3} {-} \, 1.73X_{4} + \, 5.96X_{5} + 6.86X_{5} X_{1} {-}2.83X_{3} X_{5} \\ & \;{-} \, 1.71X_{4} X_{5} + \, 2.89X_{3}^{2} {-}2.17X_{1} X_{2} X_{5} {-}2.33X_{1} X_{3} X_{5} {-}1.09X_{1}^{2} X_{5} {-} \, 1.29X_{3}^{2} X_{5} \\ \end{aligned}$$
(3)
To assess the adequacy and significance of the process, Analysis of Variance (ANOVA), F-value, and p-value were utilized (Table 3). The analysis of variance results (Tables) showed a strong agreement between the experimental and predicted values. The ANOVA results showed a Model F-value of 29.85 and P-values lower than 0.05, indicating the significance of the model terms. The findings indicated that the quadratic model is a suitable model and showed high significance with a confidence level of 93% (F = 29.58; p < 0.0001). The predicted R2 is 0.8268, which reasonably aligns with the adjusted R2 of 0.9025, showing a difference of less than 0.2. The Adeq Precision, assesses the signal-to-noise ratio, which exceeds the desired ratio of 4, with a value of 29.158, indicating a satisfactory signal. The analysis of variance in Table 3 showed that the lack of fit of the model was not significant. Therefore, this model is reliable for navigating the design space.
Table 3 Analysis of variance (ANOVA) for response surface quadratic model (Y).Investigation of the accuracy and validity of the proposed modelThe accuracy and validity of the proposed model for the removal of TBBPA were evaluated by examining the normal plot of residuals predicted versus. In Fig. 8A, the experimental data points align closely with a straight line, confirming the reliability of the model in predicting TBBPA removal. In Fig. 8B, the model shows that the normal plot of residuals is adequate. In Fig. 8C, Box-Cox analysis determined an optimal lambda value of 0.8, which is equivalent to 1 in this model, further validating the model fit. These results also support the choice of the quadratic model, demonstrating its suitability for the experimental data on response variables.Figure 8Experimental data versus predicted data by the statistical model (A) normal probability of residuals (B) Box-Cox plot (C).Determination of optimal conditionThe optimal conditions for achieving maximum TBBPA removal efficiency using Fe–S/CuS nanocomposite indicate that the highest adsorption efficiency reached 97.76% under ideal conditions. These conditions include a contact time of 15 min, pH of 7, Fe–S/CuS nanocomposite dosa ge of 0.69 g L−1, salt concentration of 0.1%, and TBBPA concentration of 15 mg L−1.Effects of pHThe 3D response surface plots in Fig. 9 show the TBBPA removal efficiency using Fe–S/CuS nanocomposite based on pH, contact time, adsorbent dose, and salt concentration. pH significantly influences the stability of TBBPA, along with the surface charge and ionization state of the adsorbent44. The efficiency of TBBPA removal at a neutral pH with a concentration of 15 mg L−1 using Fe–S nanoparticle was 58.54% within 15 min. In comparison, the Fe–S/CuS nanocomposite exhibited a significantly higher removal efficiency of 97.28% under the same conditions. The point of zero charge (pHzpc) for Fe–S/CuS nanocomposite was determined to be approximately 5. For conditions where the pH exceeds the pHzpc (pH = 5–9), the adsorbent surface becomes negatively charged. TBBPA, a hydrophobic ionizable organic compound, has two acidic hydrogens with pKa values of 7.5 and 8.544. At a pH > 7.5, TBBPA undergoes dissociation and primarily exists in the form of a negatively charged phenoxy ion45. Moreover, the surface electric potential and functional groups on the surface of the Fe–S/CuS nanocomposite deprotonated have negatively charged. Therefore, the electrostatic repulsion between the Fe–S/CuS nanocomposite surface with a negative charge and charged phenoxy ion in the solution leads to a decrease in the adsorption capacity44. At pH < 7.5 to acidic conditions, due to the protonated TBBPA and the negatively charged surface of Fe–S/CuS nanocomposite in the solution from 5 to 7.5, probably adsorption capacity occurred due to the electrostatic attraction between TBBPA and Fe–S/CuS nanocomposite (Fig. 9A)45.The increase in surface acidity is anticipated to enhance the interaction between the electrons of the aromatic ring and the surface, enhancing sorption capacity. Consequently, within the pH range (pH = 5–7.5), TBBPA gains increased and easy access to the active sites on the surface of the adsorbent. In a highly acidic solution (pH = 3), the water solubility of TBBPA was very limited. Kang et al. reported that the efficiency of TBBPA removal by Au/Fe@biocarbon was achieved within 100 min under both acidic and neutral conditions (pH = 3–7). This was due to electrostatic repulsion between Au/Fe@biocarbon and TBBPA, confirming the observed results46. Fasfous et al.’s study on TBBPA removal using MWCNTs found that the highest removal efficiency occurred at a pH of 7. The efficiency was significantly enhanced by increasing the pH from 5 to 7.5. However, beyond a pH of 7.5, the results showed a reversal in the trend44.Figure 9Removal efficiency of TBBPA using Fe–S/CuS nanocomposite as a function of (A) pH and contact time [TBBPA concentration 15 mg L−1; adsorbent dose 0.69 g L−1], (B) adsorbent dose and contact time [TBBPA concentration 15 mg L−1; solution pH 7, adsorbent dose 0.69 g L−1], and (C) Salt concentration and contact time [TBBPA concentration 15 mg L−1 solution pH 7; adsorbent dose 0.69 g L−1].Adsorbent doseThe results indicated that the removal efficiency was highest at the lowest adsorbent dose (Fig. 9B). Due to its magnetic and cumulative properties, as the amount of adsorbent increased, the adsorbent particles stuck together and overlapped each other47. Therefore, the adsorption efficacy decreases with an increasing adsorbent dose. In the study conducted by Sobhanikia et al., it was found that increasing the dosage of zero iron nanoparticles from 0.2 g L−1 led to a slight decrease in the removal of antibiotic penicillin G. This decrease in removal was attributed to the accumulation and adhesion of zero iron nanoparticles at higher doses, resulting in the loss of active absorbent sites48.Ionic strength (NaCl)It was observed that the removal efficiency of TBBPA gradually decreased with the increase in salt concentration (Fig. 9C). NaCl may compete with TBBPA for adsorption sites on Fe–S/CuS nanocomposite49. In addition, the decrease in absorption rate can be attributed to the small size of the NaCl molecule, which occupies some of the absorbent sites before the pollutant, preventing the absorption of TBBPA on the absorber50,51. Moreover, as the NaCl concentration increased, the salinity of the solution also increased, leading to a reduction in the mass transfer rate52. Salt neutralizes the absorbent surface, reducing its absorption rate and load53.In the study by Jie et al., it was shown that the removal of TBBPA was minimally affected by the salt concentration in ozone oxidation performance52. In the study by Arbeli et al., it was found that an increase in NaCl concentration in sediment led to a reduction in the removal of TBBPA through dehalogenation54.Contact timeContact time is one of the most important and influential parameters for adsorption efficiency in certain situations and for economic reasons44. Results showed that TBBPA absorption was highest in the first few minutes, with only a slight increase when extending the contact time to 90 min. High absorption occurred rapidly, surpassing 90% within 15 min and reaching equilibrium by 50 min, with minimal changes thereafter. This could be attributed to the abundance of active sites on the Fe–S/CuS nanocomposite adsorbent. Over time, these active sites became occupied by TBBPA molecules. A study by Zhou et al. on using Fe3O4@polyaniline for TBBPA removal showed that increasing contact time led to higher removal efficiency, with the sorption process following two steps49.The kinetic studiesThree kinetic models, including pseudo-first-order kinetics, pseudo-second-order kinetics, and the intraparticle diffusion model, were used to analyze the experimental data and assess the adsorption process. To elucidate the adsorption kinetic, pH of 7, FeS–CuS nanocomposite dosage of 0.69 g L−1, salt concentration of 0.1%, and TBBPA concentration of 15 mg L−1 were applied. Table 4 provides the results for the three kinetic models, including their R2 values, linear form, model equations, parameters, and reaction rate constants. Figure 10 shows the kinetic plots for the removal of TBBPA by the Fe–S/CuS nanocomposite. For the three kinetic models at varying concentrations, the calculated sorption capacities, reaction rate constants and R2 values of the linear correlation coefficients are shown in Table 2.
Table 4 Parameters of kinetic models fitted to TBBPA adsorption experiments.Figure 10Kinetic plots for the removal of TBBPA by the Fe–S/CuS nanocomposite.The pseudo-second-order adsorption model was selected for this process due to its high correlation coefficient (R2 > 0.99), suggesting a strong fit for the adsorption kinetics of TBBPA onto the Fe–S/CuS nanocomposite surface30. In the study by Zhou et al., the adsorption kinetics of TBBPA by MCNTs@ZIF-67 were well-fitted to a pseudo-second-order kinetic model55.Initial TBBPA concentrationIn this study, the initial concentrations considered were in the range of (5, 10, 15, 20, and 30 mg L−1). As the concentration of TBBPA increases, the removal efficiency decreases. This is due to the limited number of active adsorption sites on the adsorbent surface. When a higher number of contaminant molecules are present in the reaction, it can hinder the effectiveness of the process, especially when the pollutant load is high. This can lead to a decrease in removal efficiency30,34. Increasing the amount of pollutants during the reaction may result in the production of more intermediaries. These intermediaries can compete with TBBPA for active adsorbent locations, ultimately reducing the adsorption rate56. Yu et al. and Li et al. have both report that increasing the initial concentration of TBBPA decreases the removal efficiency57. Additionally, Zhang et al. found that as the initial concentration of TBBPA increased from 5 to 20 mg L−1, the rate of TBBPA removal by MnO2/MWCNT-Ni Composite decreased56.Isotherm studiesIn this study, particular emphasis was placed on simulating the most significant isotherms among several, with the Freundlich and Langmuir models being employed. The experimental data and corresponding models are presented in Fig. 11 and Table 5. The R2 value obtained from the Freundlich model (R2 = 0.9964) exceeded that of the Langmuir model (R2 = 0.9596) indicating that the Freundlich model provides a better description of the adsorption of TBBPA onto the Fe–S/CuS nanocomposite. Freundlich demonstrated both multi-layer sorption and monolayer sorption, assuming that the adsorption occurs on a heterogeneous surface with a multilayer adsorption mechanism58. In the Freundlich model, the parameter “n” represents the surface heterogeneity factor. In the present study, the values of n > 1, signify favorable adsorption and an increase in adsorption capacity with a higher initial concentration of TBBPA. Isotherm studies conducted by Zhou et al. on TBBPA removal with magnetic core–shell Fe3O4@polyaniline showed that the adsorption process, similar to our study, followed the Freundlich model49.Figure 11Absorption isothermal graph fraundlich at different TBBPA concentrations.Table 5 Isotherm parameters for the adsorption of TBBPA onto the Fe–S/CuS nanocomposites.Potential reaction mechanisms that affect the removal efficiency of TBBPA include both physical and chemical processes45. Since the solubility of TBBPA is low, one of the mechanisms involved is hydrophobicity17. The active sites on the Fe–S/CuS nanocomposite surface interact with TBBPA molecules in the solution through charged phenoxy ions44. Furthermore, the electrostatic capacity of the Fe–S/CuS nanocomposite enhances the interaction with the aromatic TBBPA, specifically targeting the double bonds on the benzene rings17 .In summary, the efficiency of TBBPA removal depends on the relative importance of factors such as the electron transfer capacity of Fe–S and CuS, surface area, and hydrophobicity17,59.Real sampleAfter preparing and analyzing the sample, the HPLC chromatogram of the real sample revealed a small peak for TBBPA. Subsequently, a concentration of 15 mg L−1 of TBBPA was added to the pre-prepared sample. Figure 12C shows the HPLC chromatogram before the treatment adsorption process. Figure 12D shows the real sample spiked with 15 mg L−1 of TBBPA after the adsorption process. This demonstrates the Fe–S/CuS nanocomposite’s ability to efficiently remove 15 mg L−1 of TBBPA, along with some of its decomposed components.Figure 12HPLC TBBPA in samples: (A) real sample; (B) not adsorb pure sample (C) real sample spiked with 15 mg L−1of TBBPA before the adsorb process; (D) real sample spiked with 15 mg L−1of TBBPA after the adsorb process (adsorbent dose = 0.69 g L−1, PH = 7, salt concentration = 0.1%).
Enhanced removal of tetrabromobisphenolA (TBBPA) from e-waste by Fe–S nanoparticles and Fe–S/CuS nanocomposite with response surface methodology (RSM)
