Magnetic Fe3O4@Mg/Al-layered double hydroxide adsorbent for preconcentration of trace metals in water matrices

A magnetic Fe3O4@MgAl-layered double hydroxide (MLDH) nanocomposite was successfully synthesized and applied as an effective adsorbent for preconcentration of trace As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions from complex matrices. The quantification of the analytes was achieved using the inductively coupled plasma optical emission spectrometry (ICP-OES) technique. The nanocomposite was then characterized using BET, FTIR, SEM, and EDS. Due to its high adsorption surface area, compared to traditional metal oxide-based adsorbents, MLDH nanocomposite exhibited high extraction efficiency. Several experimental parameters controlling the preconcentration of the trace metals were optimized using response surface methodology based on central composite design. Under optimum conditions, the linearity ranged from 0.1 to 500 µg L−1 and the correlation of coefficients (R2) were higher than 0.999. The limits of detection (LODs) and quantification (LOQs) were 0.11–0.22 µg L−1 and 0.35–0.73 µg L−1, respectively. The intra-day (n = 10) and inter-day precisions (n = 5 working days) expressed in the form of percent relative standard deviations (%RSDs) were below 5%. The proposed method was successfully applied for the analysis of the As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions in different environmental water samples.


Instrumentation. Inductively coupled plasma-optical emission spectrometer (ICP-OES) (iCAP 6500
Duo, Thermo Scientific, UK) equipped with a charge injection device (CID) was employed for quantification of As(III), Cd(II), Cr(III), Co(II), Ni(II) and Pb(II) ions. The ultrasound-assisted preconcentration of trace metals was performed in a Scientech ultrasonic bath system with a frequency of 50 Hz and power of 150 W (Labotec, Midrand, South Africa). Scanning electron microscope (SEM) (JSM-6360LVSEM, JEOL Co., Japan) was used to examine the morphology Fe 3 O 4 @Mg/Al-LDH nanocomposite. The X-ray powder diffraction (XRD) patterns recorded using a Philips X-ray generator model PW 3710/31 a diffractometer with an automatic sample changer model PW 1775. The specific surface area, pore size, and pore volume values were measured using Surface Area and Porosity Analyzer (ASAP2020 V3. 00H, Micromeritics Instrument Corporation, Norcross, USA). Spectrum 100 FT-IR (PerkinElmer, USA) spectrometer equipped with Universal Attenuated Total Reflectance (ATR) was used to investigate the functional groups on the surface of the adsorbent, X-ray photon spectroscopy (XPS) on VG ESCALAB MARK II (VG, UK).
Sample collection and preparation. The sampling of real water samples (tap and river water) was collected from south of Johannesburg, South Africa, and were kept in 1000 mL polypropylene bottles and stored in the fridge at 4 °C until analysis. Prior to analysis, the water samples were filtered using 0. 25  Whilst continuing stirring, 25 mL aliquot of ammonium hydroxide solution was added dropwise into the solution to form a black precipitate. The mixture could stir at the same conditions for 15 min and then cooled at ambient temperature. The Fe 3 O 4 was collected using an external magnet. The magnetic nanoparticles were rinsed with deionized water. The magnetitic nanoparticles were dried at 60 °C for 10 h in an over and then they were pulverized into fine particles using a pestle and mortar. Preconcentration of trace metals using UA-DMSPE method. The ultrasonic-assisted dispersive solid-phase extraction procedure was carried according to our previous methods 12,32 . Briefly, 50-100 mg of the magnetic LDHs nanocomposite was weighed into 100 mL plastic bottles, then an aliquot of 20 mL synthetic sample solutions (pH 3-9) was added to the sample bottles. The extraction and preconcentration of the As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions were carried out by ultrasonication for 5-30 min. After the preconcentration process, the sample solution and nano adsorbent were separated using an external magnetic. The analytes desorbed by adding 2.5 mL of 0.1-1.0 HNO 3 to the adsorbent and sonicated for 5 min. The eluent and the nanoadsorbent were separated by magnetic decantation. The aqueous phase containing the analytes was filtered into a pre-cleaned polypropylene centrifuge tube using the PVDF syringe membrane. The filtrate was then analysed by ICP-OES. The experimental parameters affecting the preconcentration method were optimized using RSM employing a CCD. These parameters include the mass of adsorbent (MA), extraction time (ET), sample pH; and eluent concentration (EC). Under optimal conditions, the above method was repeated for the validation and application of the proposed method.

Synthesis of
Method validation for the determination of trace metals. The performance and validation of the developed method were assessed using the linear range, precision (repeatability and reproducibility), preconcentration factor (PF), enhancement factor (EF), accuracy, the limit of detection (LOD), and limit of quantification (LOQ). The accuracy and precision (intraday and interday) were investigated using certified reference materials (ERM ® -CA713) and standard reference materials (NIST SRM 1640a and 1643e). The accuracy was expressed as percentage recovery (%R) and relative error (%RE) and they were calculated according to Eqs. (1) and (2). The precision was expressed as relative standard deviation (%RSD) which was calculated using Eq. (3), where S d is the standard deviation of five and ten replicates of the CRM analysis.
The LOD and LOQ were calculated as 3S d m and 10S d m , where S d is the Standard deviation of ten measurements of blank solution (analysed and m is the slope of the respective investigated trace metal ion calibration curve. The linearity was investigated by preconcentrating a series of standard solutions ranging from 0 to 500 µg L −1 .

Results and discussion
Characterization of the nanocomposite. X-ray diffraction (XRD). Figure 1 presents the X-ray dif- (1) %R = Obtained value Certified value × 100  31,35 . Similar peaks can be observed for pristine Fe 3 O 4 nanoparticles. These findings proved that the prepared adsorbent was crystalline and they were similar to those reported elsewhere 35 .
Morphological properties and elemental composition of the nanocomposite. The prepared magnetic Fe 3 O 4 @ MgAl LDH was characterized by scanning electron microscope/(SEM/EDS) and transmission electron microscopy (TEM) to examine the morphological properties as well as elemental composition ( Fig. 2a,b). The Fe 3 O 4 @ MgAl LDH nanocomposite shows various sizes of particles which suggests that the adsorbent has several sorption active sites that can lead to higher adsorption capacities 30 . The EDS spectrum (Fig. 2c) shows that the nanocomposite was composed of Fe(31.58%), Mg(0.97%), Al(3.14%), and O(38.80%), confirming the successful synthesis of Fe 3 O 4 @MgAl LDH adsorbent and C(25.20%) which was from the carbon coating.
Brunauer-Emmett-teller (BET). Nitrogen adsorption/desorption isotherm and the subsequent pore size distribution (obtained from the adsorption/desorption isotherm) for the Fe 3 O 4 @MgAl LDH composite are illustrated in Fig. 3. As shown in Fig. 3a, the curve has a distinct hysteresis loop that was observed in the pressure range of 0.4-1.0 p/p 0 which is a typical type IV isotherm. This implied that the prepared composite was mesoporous. The BET specific surface area pore volume and pore size ( Fig. 3b- was successfully coated with MgAl LDH. These results are comparable to the literature 7,24,33 and they were better than those reported elsewhere 7,31 . The relatively high specific surface area proved that the prepared adsorbent is suitable for the adsorption of trace metals. Fourier-transform infrared spectroscopy (FTIR). The Fourier-transform infrared spectroscopy (FTIR) technique was used to assess the functional groups present on the surface of Fe 3 O 4 @MgAl-LDH nanocomposite and the results are presented in Fig. 4. The broad peak at 3434 cm −1 was ascribed to O-H stretching of LDH layers and interlayer water molecules 7,33,36 . A sharp at 1374 cm −1 was assigned to the stretching vibration of intercalated Selection of the adsorbent. The selection of a suitable and effective adsorbent is one of the critical aspects of the extraction and preconcentration of trace metals. Therefore, in this study, the sorption and extraction ability of Optimization of the preconcentration procedure using a central composite design. The CCD matrix and respective recoveries (analytical response) for each analyte are presented in supplementary data (Table S1). Analysis of variance (ANOVA) was carried to investigate the quality of the RSM model and assess the most significant parameters as well as to examine the interactions between the independent variables. The ANOVA results reproduced in the form of a Pareto chart (Fig. 6) revealed that the p-values of the individual variables were less than 0.05, implying that they are significant at a 95% confidence level.  www.nature.com/scientificreports/ Response surface methodology. The three-dimensional response surface plots of the second-order polynomial equation were used to examine interactions between independent variables and their combined effect on the analytical response. The combined effect of two variables was assessed by varying them within the determined ranges while keeping two other variables at zero level (central point) (Fig. 7). The response surface plots for the effect of pH and eluent concentration, pH and extraction time, and pH and mass of adsorbent on the preconcentration of target analytes show that the % recovery increased with a slight increase in sample pH. These observations agreed with the Pareto chart illustration which demonstrated that sample pH was ranked as the least significant factor. The response surface plots showing the effect of the mass of adsorbent and pH, the mass of adsorbent and eluent concentration as well as the mass of adsorbent and extraction time revealed that the analytical response for the target analytes increased with an increase in the mass of adsorbent. This might be due to the availability of adsorption sites on the surface of the adsorbent which enhances the extraction efficiency of the adsorbent. Lastly, the increase in extraction time and eluent concentration also proved to have a positive effect on the analytical response.
Optimization by desirability function. The desirability function was employed to optimize individual variables (sample pH, extraction time, eluent concentration, and mass of adsorbent) that affect the pre-concentration process. The desirability values of 0.0, 0.5, and 1.0 corresponded to the minimum, middle and maximum functions. The values in the desirability profile (Fig. 8) that are closer to 1.0 imply that the corresponding parameter or variable condition is optimum 37 . The optimum conditions for simultaneous preconcentration of As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions were selected to be 6.5, 84 mg, 10 min, and 2.0 mol L −1 for sample pH, mass of adsorbent, extraction time and eluent concentration, respectively.
Adsorption capacity. The adsorption capacity of the nanocomposite was investigated under optimum con- ditions. An aliquot of 100 mL synthetic solutions containing 10-250 mg L −1 of As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions was processed using the optimized preconcentration method. After equilibrium has been reached the solutions were determined using ICP-OES. The adsorption capacity was measured by using the following formula:    Table 1.
where q e : amount adsorbed; q max : maximum monolayer adsorption; K L : Langmuir constant; C e ; concentration of adsorbate at equilibrium; RL: separation factor; K F : adsorption capacity; 1/n: adsorption intensity. The adsorption isotherms of the Fe 3 O 4 @MgAl LDH for As, Cd, Co, Cr, Ni, and Pb were performed under the optimized conditions. The results are shown in Table 1. As reported in Table 1, the R 2 and RSE values of Langmuir models were higher than the R 2 and RSE values of Freundlich. This suggested that the adsorption processes conformed well to the Langmuir model. The experimental maximum adsorption capacities for As, Cd, Co, Cr, Ni, and Pb Langmuir isotherms were found to be (124, 122, 138, 116, 120, 130) and for Freundlich were 0, 24.6, 22.6, 21.6, 23.5 and 25.2).
Adsorption mechanism. Even though the FTIR spectra of Fe 3 O 4 /MgAl LDH nanocomposite before and after adsorption of As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) (Fig. 9a) showed no noticeable changes. The XRD patterns, EDS, and TEM spectra respectively were used for clarifying the removal mechanisms ( Fig. 9b-d). It can be observed from XRD patterns demonstrated by (Fig. 9b) showed high crystallinity of the compounds before and after adsorption. The typical diffraction peaks of the Fe 3 O 4 @MgAl LDH before adsorption and after adsorption were still strong and sharp. The EDX of Fe 3 O 4 @MgAl LDH nanocomposite after adsorption shows (Fig. 9d,e) it is mainly composed of the elements O, Fe, Mg, Al, As, Ni, Cd, Cr, Co, and Pb this confirms the adsorption of the analytes by the adsorbent was successful.
The survey spectra of Fe 3 O 4 @MgAl LDH nanocomposites confirm the existence of Al, Mg, Fe, O, and C atoms on the nanocomposites before and after adsorption. Noticeably, the survey spectra of Fe 3 O 4 @MgAl LDH nanocomposite after the adsorption of metal ions show an increase in the intensity signal as compared to before adsorption, thus suggesting the conceivable adsorption of metal ions in solution (Fig. 10a) 38 . Subsequently, the high-resolution XPS spectra of Fe2p (Fig. 10d) exhibited two main peaks at 711.25 and 724.41 eV corresponding to Fe2p3/2 and Fe2p1/2, respectively. The deconvoluted Fe2p spectra before and after adsorption displayed Fe 2+ and Fe 3+ at 710.24 and 712.27 eV confirming the existence of Fe 3 O 4 in the nanocomposite 19 . No significant shifts in the peaks on Fe2p were observed considering the binding energy before and after adsorption, indicating that Fe 3 O 4 retained its chemical structure on the composite. As expected, the high-resolution spectra of Al2p (Fig. 10b) and Mg1s (Fig. 10e) with binding energies of 73.8 and 1304.2 eV show negligible changes before and after adsorption.
The deconvoluted high-resolution spectra of C1s (Fig. 10c) shows three-carbon contributions corresponding to the C-C, C-O and O-C=O at binding energies of 284.5, 286.1 and 288.5, respectively, before and after adsorption. Subsequently, the C1s signal increased after adsorption from 22.5 to 23.9%. The increase in the C1s signal can be subjected to the increase in interlayer CO 3 2− anions 39 . Moreover, the increase in the C1s signal correlates with the reduction in the O1s signal, as the reduction on the surface hydroxyl groups was complemented by the increase on the intercalated CO3 2− anions, which is one of the adsorption sites that can effectively adsorb metal cations of interest [39][40][41] . Interestingly, the Ols (Fig. 10f) content decreased after adsorption, indicating possible adsorption of As, Pb, Cd, Cr, Co, and Ni metals through the interaction with the present hydroxyl groups on the surface of the composite, which corroborates with the EDS results. Although adsorption primarily occurs through the interaction of the analyte with the surface hydroxyl groups, another site of adsorption concerning the intercalated CO 3 2− has been reported 40 . This implies the adsorption of As, Pb, Cd, Cr, Co, and Ni occurs on both the intercalated MgAl-CO 3 2− or on the surface MgAl-OH x /deprotonated MgAl-O −40 . To this end, the possible mechanisms of adsorption related to these interactions are listed as follows: www.nature.com/scientificreports/

Analytical figures of merit and method validation.
Analytical figures of merit of the UASPME/ICP-OES method, such as linear range, correlation coefficient (R 2 ), preconcentration factor (PF), enhancement factor (EF), precision (intraday and interday), LODs, and LOQs were determined as discussed in "Experimental" section for the target analytes. The calibration curve equations, linearity, R 2 , and the enrichment factors for each analyte are presented in Table 2.
The preconcentration factor, calculated as the ratio of the sample volume (100 mL) divided by the eluent volume (2.5 mL), was found to be 40. The LODs and LOQs of the UASPME/ICP-OES method are illustrated in Table 3. The precision (expressed (%RSD)) of the method was investigated in terms of intraday (repeatability, n = 10) and interday (reproducibility n = 5 working day). The %RSD for the analysis of As, Cd, Co, Cr, Ni, and Pb by the UASPME/ICP-OES method were less than 5% (Table 3). Table 4 displays the results obtained for the determination of As(III), Cd(II), Cr(III), Co(II), Ni(II) and Pb(II) from NIST SRM 1643e (trace elements in water), CRM ERM-CA713 (trace metals in wastewater) and NIST SRM 1640a (trace elements in natural water). The relative error (%RE) values ranged from − 2.4 to 1.7%, signifying that the developed method had high accuracy. Additionally, the obtained values agreed with the certified values at a 95% confidence level.
Additionally, the validity and applicability of the developed UASPME/ICP-OES procedure were evaluated by analysis of As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions in complex matrices such tap and river water. The accuracy of the method was evaluated by spiking the water samples with 10 µg L −1 of the target analytes the analytical results are presented in Table 5. As can be seen, the recoveries for the trace elements in each type of water sample were above 95% standard deviations less than 3%.  Table 6. These findings suggested that the proposed method had the capability to extract, preconcentrate, and determine trace metals ions in real samples with complex matrices. The comparison of analytical characteristics of UASPME/ICP-OES method for preconcentration of As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions with previously researched preconcentration procedures specifically (solid-phase based extraction) reported in the literature are summarised in Table 7. It was observed that the proposed method had superior LOD, EF, RSD, and linear range compared to those reported elsewhere 42,43 . In the addition, parameters such as liner calibration range, LOD, EF, and RSD for UASPME/ICP-OES technique were www.nature.com/scientificreports/ Table 2. Calibration curve equations, correlation coefficients and enrichment factors of UASPME/ICP-OES method for As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions. a Enrichment factor was defined as the ratio between the slope of the calibration curves before and after preconcentration.  Table 3. Analytical performances of UA-DSPME/ICP-OES method.   Figure 11. illustrate the % recoveries of each investigated metal ion. As shown in Figure,

Conclusion
The Fe 3 O 4 @MgAl LDH composite was prepared successfully via the co-precipitation method. The prepared composite was characterized through instruments such as SEM-EDS, TEM, BET surface area, and FTIR. The Fe 3 O 4 @MgAl LDH composite was applied as an adsorbent in UA-DSPME for preconcentration of trace As(III), Cd(II), Cr(III), Co(II), Ni(II), and Pb(II) ions in river water, tap water, and mine wastewater samples. Experimental parameters affecting UA-DSPME/ICP-OES method were optimized using a multivariate approach. The UA-DSPME/ICP-OES method revealed outstanding analytical characteristics such as simplicity, rapidity, low LODs, high accuracy, enrichment factors, and precision. Furthermore, the performance of the UA-DSPME/ ICP-OES method was applied for the determination of trace metals in complex matrices. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.