Introduction

Recently, access to clean water has become a global concern. The remarkable increase in human population has triggered high industrialization and agricultural production to meet mankind sustainability (Munonde 2017). These activities lead to increased use of various chemical substances that are then released into the environment. Among these chemical substances, neonicotinoid insecticides (also known as neonics) are the mostly distributed into the environment via agricultural practices (Vichapong et al. 2015). Neonicotinoid insecticides are a relatively new class of insecticides derived from nicotine, and they were introduced in the 1990s for insects control (Yáñez et al. 2014; Farajzadeh et al. 2016). They act agonistically at the nicotinic acetylcholine receptors (nAChRs) on the post-synaptic layer. These receptors contribute greatly to signal transmission in the central nervous system (CNS) (Vichapong et al. 2013; Jovanov et al. 2013; Zhao et al. 2020; Guo et al. 2020; Yamamoto and Casida 1999). Neonicotinoids bind tightly to the nAChRs of insects than to any other organism including mammals, and the latter can cause paralysis or even death. However, these compounds are considered less toxic to mammals (Farajzadeh et al. 2016; Zou et al. 2019; Mahai et al. 2019).

Several nations have drafted the allowable residue limits of these compounds in various environmental matrices (Zhao et al. 2020; Pietrzak et al. 2020). Moreover, European Union (EU) legislation has established the strict regulations for the maximum consumable residue limits of the neonicotinoids in various agricultural products and water. The ubiquity of these compounds in the environmental waters is of great global concern, and according to the EU legislation, the total maximum permissible levels for neonicotinoids and their metabolites should be 1.05 µg L−1 (Anzecc 2000; Yi et al. 2019). Each neonicotinoid insecticide can produce various metabolites depending on the environmental conditions. It is worth noting that the current work focuses on the actual neonicotinoids rather than the metabolites. Neonicotinoids can in turn become toxic to humans and other animals if they are continuously consumed in amounts exceeding maximum residue limits (MRLs). Therefore, proper monitoring and removal of these compounds are necessary to promote the safety of terrestrial and aquatic organisms (Sánchez-Bayo and Hyne 2014).

High-performance liquid chromatography (HPLC) coupled with different suitable detectors including fluorescence (Treder et al. 2020), diode array detector (Arase et al. 2016; Zhou et al. 2006), and tandem mass spectrometry (Ike et al. 2018; Kachangoon et al. 2020) have attracted a lot of attention for the analysis of organic pollutants including neonicotinoids owing to its superior sensitivity and reliability (Song et al. 2018). However, the direct analysis of the neonicotinoids by HPLC is difficult due to their low concentrations in complex environmental matrices (Arnnok et al. 2017; Azzouz et al. 2018). Therefore, a sample pre-treatment technique is needed prior to their quantitative analysis. Recently, numerous sample pre-treatment techniques have been reported (Guo et al. 2020). To date, researchers are still using traditional sample preparation methods such as solid-phase extraction (SPE) (Furman et al. 2010; Mpupa et al. 2019; Georgakopoulos et al. 2011) and liquid-phase extraction (LLE) (Jovanov et al. 2013; Peyton 1993; Rodríguez et al. 2016). However, these techniques are tedious, laborious, expensive and environmentally unfriendly. In addition, LLE requires large volumes of toxic organic solvents (Zou et al. 2019). Therefore, new methods based on the miniaturization of the traditional procedures, improvement of selectivity to minimize matrix effects and reduction in sample pre-treatment steps and time are necessary (Ike et al. 2018; Kachangoon et al. 2020). To mitigate the aforementioned setbacks, microextraction techniques that have been developed include solid-phase microextraction (SPME) (Zhao et al. 2015; Jiang et al. 2015), liquid-phase microextraction (LPME), dispersive solid-phase microextraction (d-SPME), dispersive liquid-phase microextraction (d-LPME) (Javadi et al. 2020) and cloud point extraction (CPE) (Ghouas et al. 2016), among others. These techniques can selectively preconcentrate the analytes of interest and subsequently clean-up the sample matrix with greater efficiency. The current work applies d-SPME using magnetic nanoparticles referred to as magnetic dispersive solid-phase microextraction (d-MSPME) which has been reported in the literature (Mpupa et al. 2019; Georgakopoulos et al. 2011; Spietelun et al. 2013; Giakisikli and Anthemidis 2013). This method offers great advantages such as the use of small organic solvents volumes, short extraction times, simpler operation and facile sorbent regeneration (Mehdinia and Aziz-Zanjani 2013).

Various nanostructured magnetic-based materials have been utilized in sample preparation methods owing to their unique physical and chemical properties. The super-paramagnetism properties of these materials enable them to be easily separated by the external (Biswas et al. 2020; Zhang et al. 2019). Additionally, the magnetic materials containing the pollutants can be magnetically recovered from the sample matrix by applying external magnet (Giakisikli and Anthemidis 2013; Vu et al. 2017). Recently, carbon-based materials have been extensively explored in sample preparation methods owing to their high adsorption capacities and extraction recoveries for organic pollutants (Chisvert et al. 2019; Zhang et al. 2013; Ebrahimi et al. 2019). Among these materials, activated carbon has shown interesting features such as high surface area, abundant surface functionalities, meso- and micro-porosity, abundant precursors and affinity to organic pollutants (Mui et al. 2010; Mogolodi Dimpe and Nomngongo 2019; Albishri et al. 2017). However, this material has poor selectivity towards a wide range of organic pollutants. Additionally, activated carbon lacks dispersibility.

In this work, poly (3 aminobenzoic acid) (PABA) was introduced to improve the properties of activated carbon. Poly (3 aminobenzoic acid) is a hydrophilic, conductive and chemically stable polymer, and it has high dispersibility characteristics (Ramohlola et al. 2017a). The incorporation of PABA into AC can strengthen various factors by tuning the morphologies and structural properties of the parent material (Ramohlola et al. 2017b). To the best of our knowledge, this is the first study reporting the application of magnetic PABA-based AC for the preconcentration of neonicotinoids in real water samples. Four neonicotinoids that are mostly found in environmental wastewater, acetamiprid (ACT), imidacloprid (IMD), thiacloprid (TCL) and clothianidin (CLD), were selected as the analytes. The preconcentration technique was revealed to be satisfactory for the simultaneous determination of these analytes.

Experimental

Materials, reagents, and standards

Acetamiprid (ACT), imidacloprid (IMD), thiacloprid (TCL) and clothianidin (CLD), 3-aminobenzoic acid, ammonium persulfate (APS), iron chloride tetrahydrate, iron chloride hexahydrate, acetonitrile (ACS reagent and HPLC grade), hydrochloric acid (37%), sodium hydroxide, ammonium solution (25%) and ethanol were purchased from Sigma-Aldrich (St. Louis, MO, USA). Tyre-based activated carbon was obtained by thermal treatment and subsequent H2O2 treatment for activation (Dimpe et al. 2018). A stock solution (100 mg/L) containing a mixture of ACT, IMD, TCL and CLD was prepared by accurately weighing each analyte, dissolving them in 100-mL acetonitrile, and the resulting solution was kept at 4 °C. Working solutions were prepared daily by diluting appropriate volumes of the original stock solution with ultrapure water produced by Direct-Q® 3UV-R purifier system, (Millipore, Merk, Germany), and ultrapure water was used throughout the experiments.

Instrumentation

Chromatographic analysis was carried out on a HPLC Agilent 1200 Infinity series, equipped with a diode array detector (Agilent Technologies, Waldbronm, Germany). The chromatograms were recorded at 268 nm, 246 nm, 272 nm, and 242 nm for CLD, ACT, IMD and TCL, respectively. An Agilent Zorbax Eclipse Plus C18 column (3.5 µm × 150 mm × 4.6 mm) (Agilent, Newport, CA, USA) was operated at a column temperature of 25 °C. The mobile phase composition contained a mixture of 70% water (mobile phase A) and 30% acetonitrile (mobile phase B), using an isocratic elution system. A flow rate of 1.0 mL/min was used throughout the analysis. The characterization of the synthesized materials was accomplished by the following analytical techniques: Infrared spectra were obtained using PerkinElmer spectrum 100 Fourier transform infrared spectrometer (Waltham, MA, USA) using a KBr disc technique. Surface area was obtained using Brunauer–Emmet–Teller surface area (SBET) by nitrogen adsorption–desorption mechanism in a Micrometric ASAP 2020, and all the gases used for analytical purposes were instrument grade. Sample pH was measured using OHAUS starter 2100 pH meter (Pine Brook, NJ, USA). Crystallinity, particle sizes and amorphous nature of the materials were discovered by Powdered X-ray diffraction (X’Pert Phillips) with CuKα radiation (0.1540 nm) polychromator beam in the 2θ scan range 20–800 was carried to elucidate the crystallinity of the material. A step time and step size of 87.63 and 0.0170 (2θ) was used at 40 kV and 40 mA instrument power settings. External morphological features were explored by scanning electron microscope (SEM, TESCAN VEGA 3 XMU, LMH instrument (Czech Republic)) coupled to energy dispersive X-ray spectroscopy (EDS) for elemental composition at an accelerating voltage of 20 kV. Internal morphologies and particles sizes were obtained by transmission electron microscopy (TEM, JEM-200X, JEOL, Japan).

Sample collection and preparation

Water samples were collected from Apies river and Daspoort wastewater treatment plant (WWTP) at Pretoria, South Africa. Daspoort is known as the wastewater plant that treat the industrial wastewater (5% of treated material) and domestic wastewater (95% of treated material). This plant uses activated sludge technology to treat wastewater. The effluent, influent and secondary wastewater were sampled and kept in glass sample bottles at 4 °C. Prior to analysis, 0.45 µm syringe filters were used to filter the samples.

Preparation of poly (3 aminobenzoic acid) (PABA) and poly (3-amino benzoic acid)/activated carbon (PABA/AC) composite

The synthesis of poly (3 amino benzoic acid) polymer was in accordance with the previously reported procedure (Ramohlola et al. 2017a). The preparation of poly (3 amino benzoic acid (PABA)-based activated carbon (PABA/AC) was by in situ chemical oxidative polymerization of m-aminobenzoic acid (3ABA) monomer along with an accurate amount of activated carbon. In short, approximately, 1 g of 3ABA monomer and 0.3 g of AC were introduced into H2O/HCl (100 mL/10 mL) solution in a 250-mL round bottom flask, and then the mixture was stirred for about 30 min at 50 °C. Subsequently, 1.88 g of FeCl3 and 2.40 g of ammonium persulfate(NH4)2S2O8 were added to the solution, and the mixture was continuously stirred for 3 h at the same temperature. The mixture was then incubated in an oven overnight at 50 °C to evaporate the solvents, and the resulting content was then filtered and dried at 50 °C.

Preparation of magnetic PABA/AC (Fe3O4@PABA/AC)

The synthesis of iron oxide nanoparticles followed the method reported previously with some modifications (Shabanian et al. 2015). The magnetic PABA/AC was obtained by following the procedure suggested by Wang et al. (2012). Briefly, an approximate amount of FeCl3.6H2O (3.03 g) and FeCl2·4H2O (1.13 g) were dissolved in 150 mL of deionized in a round bottom flask. The suspension was stirred for 30 min and the sodium dodecyl sulphate (SDS) was subsequently added, and the mixture was sonicated for 30 min to promote dispersion. The PABA/AC was then introduced into the solution mixture followed by heating up to 60 °C under N2 purging at a constant stirring. Ammonium solution (NH3OH, 25%) was then used to adjust the pH of the solution to 10 followed by heating for about 30 min under vigorous stirring. The blackish brown precipitate was then collected by external magnet and washed several times with ethanol and distilled water, and the product was dried overnight at 60 °C.

Dispersive magnetic solid-phase microextraction procedure (d-MSPME)

The model sample solution of neonicotinoid insecticides (10 µg L−1) was prepared by diluting appropriate volume of the original standard solution of 1000 µg L−1. An amount of about 10–40 mg of Fe3O4@PABA/AC composite was then introduced into the sample bottle containing 5 mL of the model solution. The extraction of neonicotinoid insecticides into the solid-phase material was achieved by ultrasonication for 5–15 min followed by the application of an external magnet to separate the supernatant from the magnetic adsorbent containing analytes. The desorption of analytes from the adsorbent was carried by the addition of 100–1200 µL of acetonitrile, and the mixture was sonicated for 5–20 min to elute the analytes from the adsorbent. The adsorbent and eluent were separated using an external magnet and the eluent was filtered using 0.22 µm PVDF syringe filters and injected into the HPLC–DAD analysis. The most influential factors were optimized using fractional factorial design (FFD) and central composite design (CCD). The levels of each factor are shown in Table 1.

Table 1 Parameters and levels applied in CCD for d-MSPME optimization
Full size table

Reusability procedure

The reusability and regeneration of Fe3O4@PABA/AC nanocomposite were studied by performing several adsorption–desorption cycles. The extraction and preconcentration experiments were conducted under optimum conditions, that is, 41 mg, 1200 µL, 10 min, 5 min and 6.5 for MA, EV, ET, DT, and pH, respectively. In each desorption process, the analyte-loaded adsorbent was mixed with 1200 µL of acetonitrile and sonicated for 5 min. The supernatant and adsorbent were separated using external magnet, and the supernatant was filtered through 0.22 µm PVDF syringe filters. The analytes in the resultant filtrate were analysed using HPLC–DAD. The spent adsorbent was washed twice with deionized water followed by ethanol and dried in vacuum oven at 60 °C for 2 h. The regenerated adsorbent was utilized for another adsorption–desorption cycle. The process was repeated several times until the adsorbent lost its adsorption capacity.

Application to real samples

The developed method was applied in real water samples for the determination of target analytes. The extraction and preconcentration procedure were performed using the optimized conditions as indicated in Sect. 2.6. Figure 1 illustrate the schematic diagram summarizing the procedure used for extraction and preconcentration of target analytes from real samples.

Fig. 1
figure 1

Schematic diagram summarizing the experimental procedures

Full size image

Results and discussion

Characterization of the synthesized materials

Fourier transform infrared spectroscopy

The chemical properties of the synthesized materials were confirmed by FTIR spectroscopy, and the spectra are shown in Fig. 2a and b. Figure 2a shows the spectra of AC, PABA and magnetic nanoparticles (magnetite), while Fig. 2b shows the spectra of PABA@AC and Fe3O4@PABA/AC nanocomposite. The FTIR spectrum of PABA shows the absorption bands at 3408 cm−1 was assigned to the overlapping stretching vibrations of carboxylic O–H and secondary amine N–H, while the bands at 1750, 1589, and 1483 cm−1 were ascribed to the stretching vibrations of C=O, COO− asymmetric stretch and −COO− symmetric stretch of the carboxylic group, respectively. The FTIR spectrum of PABA was similar to the one reported by Zare et al. (2018). The FTIR spectrum of the AC exhibited a broad peak around 3100 cm−1 which was attributed to the vibration of the surface alcoholic OH or phenolic OH groups. The incorporation of PABA on AC matrix resulted on the upward shift of the peak that was observed in FTIR spectrum of AC to s 3500 cm−1 (Fig. 2a). The AC spectrum shows absorption peaks at 1627, 1550 and 1390 cm−1 indicating the stretching vibration of C=O, the skeletal vibration of C=C and C=OH vibration, respectively (Mashile et al. 2018a). Magnetic nanoparticles spectrum showed the main intense asymmetrical peak at around 571 cm−1 corresponding to the Fe–O stretching vibrations (Tarhan et al. 2020). In Fig. 2b, the FTIR spectrum of PABA/AC exhibited a strong peak at 3448 cm−1 associated to the stretching vibration of O–H/N–H groups. The two small peaks at 2920 cm−1 and 2851 cm−1 represented the symmetric and asymmetric stretching of C–H bond (Awual et al. 2019; Naushad et al. 2019). The band at 1751 cm−1 was associated with the stretching vibration of C=O from carboxylic group. In addition, the amide groups stretching vibrations were observed at 1660 cm−1 showing the polymer bending on AC surface. The peaks at 1440 cm−1 and 1090 cm−1 indicated the stretching vibrations of N–H and C–N, respectively (Njoku et al. 2014), confirming the successful doping of PABA polymer on AC. The spectrum of Fe3O4@PABA/AC exhibited Fe–O stretching vibration peak at 571 cm−1 confirming the successful incorporation of magnetic nanoparticles in to the PABA/AC composite (Feng et al. 2020). Additionally, the disappearance of the peak at 1250 cm−1 was a sign of successful doping of iron oxide nanoparticles.

Fig. 2
figure 2

FTIR spectra indicating the functional groups and characteristic bonds of the as-synthesized materials. Magnetite refers to Fe3O4 nanoparticles

Full size image

X-ray diffraction spectroscopy

The XRD patterns for magnetite (Fe3O4 nanoparticles), AC, PABA, PABA/AC and Fe3O4@PABA/AC nanocomposite are shown in Fig. 3. The XRD pattern of PABA exhibited a broad peak at 2θ angle of 26° which is an indication of the amorphous nature of PABA polymer, and it agreed with the recently reported XRD pattern of polyaniline (Mashao et al. 2019). Magnetic nanoparticles exhibited the main diffraction peaks at 2θ angle of 30.2, 35.4, 43.2, 53.6, 57.1, 62.9 and 74.3, which corresponded with (220), (311), (400), (422), (511), (440) and (533) hkl values, respectively. Additionally, these peaks matched with the reflections proposed by Bragg based on the spinel ferrite structure (Hammad et al. 2020). The broad peaks appeared at 2θ of 10°–30° and 30°–50° were the characteristic peaks of amorphous AC, matching (022) and (100) graphitic carbon diffraction pattern (Du et al. 2020; Wang et al. 2019). In the PABA/AC composite, the broad peak of PABA was still observed indicating that the PABA polymer sat on AC clusters. In the nanocomposite (Fe3O4@PABA/AC), it was observed that the magnetic nanoparticles were immobilized on the PABA/AC composite surface. Furthermore, some of the magnetite diffraction peaks were maintained but their intensities decreased proving the successful incorporation of magnetic nanoparticles onto the surface of PABA/AC. It should be noted that there are similarities in the XRD patterns of Fe3O4 and Fe3O4@PABA/AC, and this was expected because the crystallinity of magnetite influences the morphology of the composite. However, it is worth mentioning that the XRD patterns of Fe3O4@PABA/AC had reduced peak intensities compared to XRD spectrum of magnetite indicating a successful incorporation of Fe3O4 on PABA/AC surface.

Fig. 3
figure 3

XRD patterns showing the crystallographic structure of the as-synthesized materials. Magnetite refers to Fe3O4 nanoparticles

Full size image

Brunauer–Emmet–Teller

The surface area and pores of the synthesized materials were obtained using Brunauer–Emmett–Teller (BET) technique based on N2 adsorption–desorption, and the results are presented in Table 2.

Table 2 Surface properties of the as-synthesized materials obtained by BET analysis
Full size table

It should be noted from Table 2 that the surface properties of the AC decreased dramatically after the incorporation of PABA. This could be due to the increase in the diameter of the PABA/AC which results in a decreased in surface area (Sadeghi et al. 2018). In addition, the lower amount of AC doped could also be the cause for surface properties decrease. A significant increase in surface area and pores was observed in the Fe3O4@PABA/AC composite after the incorporation of Fe3O4 nanoparticles. This could further confirm that Fe3O4 nanoparticles played a crucial role in neonicotinoids extraction by reducing the diameter of the PABA/AC composite, which in turn increases its surface area. The observed surface area and pores were enough for the maximum extraction of the neonicotinoids.

Scanning electron microscopy

Figure 4 shows the external morphologies and elemental composition of PABA/AC and Fe3O4@PABA/AC obtained using SEM/EDS. In the PABA/AC composite (Fig. 4a), PABA covered the surface of the porous AC. Furthermore, the AC clusters were minimized and were heterogeneously mixed with PABA matrix (Hasan et al. 2019). After the incorporation of Fe3O4 into the composite (Fig. 4c), pores were reduced. In addition, the coated Fe3O4 resulted in a smooth surface and reduced particles. The EDS spectrum (Fig. 4b) revealed the presence of oxygen and carbon in PABA/AC as the important elements of the nanocomposite. Chlorine and iron observed were from the oxidizing agent from PABA polymerization. After the incorporation of Fe3O4 (Fig. 4d), the weight percentage of iron increased drastically confirming the presence of Fe3O4 in the Fe3O4@PABA/AC composite.

Fig. 4
figure 4

SEM images indicating the external morphologies of PABA/AC (a), Fe3O4@PABA/AC (c) and EDS spectrum of PABA/AC (b) and Fe3O4@PABA/AC (D)

Full size image

Transmission electron microscopy

The internal morphological features of the synthesized materials were investigated using TEM analysis, and the images are shown in Fig. 5. The TEM image of PABA was similar to the previously reported image by Ramohlola et al. 2017a. In Fig. 5a, it should be observed that PABA sat on AC clusters. In the composite (Fig. 5b), Fe3O4 nanoparticles distributed dependently mostly on PABA surface owing to their affinity towards the groups that are abundant on the polymer. The agglomeration observed is caused by poor dispersibility of the Fe3O4 nanoparticles.

Fig. 5
figure 5

TEM images indicating the internal morphologies of a PABA/AC and b Fe3O4@PABA/AC

Full size image

Optimization of the developed d-MSPME

Screening of d-MSPME preconcentration procedure using central composite design

To obtain best results from any analytical method, screening process was conducted to investigate the most influential parameters in the extraction and preconcentration of neonicotinoids. A two-level (25–1) FFD was employed as a chemometric tool for the screening of d-MSPME method. In this work, five parameters including pH, extraction time (ET), mass of adsorbent (MA), eluent volume (EV) and desorption time (DT) were optimized. Twenty experiments were conducted for screening and the results were recorded as percentage recoveries (Additional file 1: Table S1). The evaluation of the experimental data was done using analysis of variance (ANOVA), and the ANOVA results were recorded as Pareto charts (Fig. 6a–d). Pareto charts allowed the assessment of the main and interactive effects. The length of each bar in Fig. 6 represented the effects of individual factors and their interactions, and these bars are proportional to the effect of each factor or interactions. If the bar length surpasses the reference line, it implies that the factor or interaction has significant effect on the analytical response at 95% confidence level (Biata et al. 2017). As shown in Fig. 6a, b and d, mass of adsorbent and eluent volume were influential on analytical response for ACT, IMD and CLD, except for extraction and preconcentration of TCL where only EV was significant at 95% confidence level. Moreover, the interaction between MA and EV (2by3) was statistically significant. Ideally, the eluent volume must be sufficient to quantitatively extract the all the analytes from the surface of the adsorbent. In addition, enough adsorbent is required to adsorb the target analytes. Therefore, both MA and EV should be sufficient to achieve quantitative adsorption and elution of the analytes, respectively. The two factors (MA and EV) were further optimized using CCD to minimize the number of experiments and to investigate their interactions based on the quadratic model employed on response surface methodologies (RSM) (Li et al. 2019).

Fig. 6
figure 6

Pareto charts indicating the effect of factors on extraction and preconcentration of a acetamiprid (ACT), b imidacloprid (IMD), c thiacloprid (TCL) and d clothianidin (CLD) using magnetic-based PABA activated carbon adsorbent. ET extraction time, MA mass of adsorbent, EV eluent volume (EV) and DT desorption time. 1by2: interaction between pH and MA; 1by3: interaction between pH and EV; 1by4: interaction between pH and ET; 1by5: interaction between pH and DT; 2by3: interaction between MA and EV; 2by4: interaction between MA and ET; 2by5: interaction between MA and DT; 3by4: interaction between EV and ET; 3by5: interaction between EV and DT; 4by5: interaction between ET and DT

Full size image

Further optimization

The preliminary studies indicated that the Fe3O4@PABA/AC was capable of adsorbing and desorbing the neonicotinoid insecticides. The factors that were affecting the performance of d-MSPME method were further investigated. These factors include MA and EV, and were optimized using CCD involving ten experiments, to evaluate their interacting effects with respect to the extraction recoveries (Additional file 1: Table S2). The interactive effects, main effects and quadratic effects towards the extraction of the neonicotinoids were investigated by the RSM based on CCD (Ferreira et al. 2019), and the results are presented as 3D surface plots in Fig. 7. The interactions between the factors were significant towards the extraction of IMD and CLD. Higher extraction recoveries were observed at higher MA and EV. However, the interactions of the factors towards the extraction of ACT and TLC did not show any statistical significance. The best precise conditions could not be assumed from the plots since the plots give the region not the precise points. Therefore, the desirability profiles were obtained by further optimization owing to their potential to pinpoint exact optimum values (Mashile et al. 2018b).

Fig. 7
figure 7

Three-dimensional contour plots of a acetamiprid (ACT), b imidacloprid (IMD), c thiacloprid (TCL) and d clothianidin (CLD) describing the combined interactive effects of investigated factors (EV eluent volume, MA mass of adsorbent) on the analytical response

Full size image

Figure 8 shows the desirability profiles of the thiacloprid obtained by the conversion of the responses into the dimensionless values called desirability by the application of statistical design. The aggregation of these individual desirability (di) are then utilized to obtain the overall desirability (Valasques et al. 2019; Bruns et al. 2006). According to the literature, desirability ranges from 0 (undesirable), 0.5 (partially desirable) and 1 (desirable) (Bezerra et al. 2019). Figure 8 also illustrates the individual desirability scores for the extraction and preconcentration of the target analytes (bottom), and the desirability value of 1.0 was selected to obtain optimum conditions. The percent recovery obtained from plots for each factor is presented at the top left-hand side. According to Mashile et al. 2018b, the figures on the top left-hand side demonstrate the changes in the level of each individual variable, analytical response, and the overall desirability. It can be seen from the desirability profile that both factors were considered significant since the lower points of these factors deviated from the desirable line indicating that a change in one of these values can cause dramatic effect on analytical responses (% extraction recoveries). The desirability profiles for the other three neonicotinoids showed similar results (Additional file 1: Fig. S1). The optimum conditions predicted from these desirability profiles and RSM were 41 mg for MA and 1200 µL for EV. The use of small MA for the extraction of the selected neonicotinoids in a short period of time has proved that the Fe3O4@PABA/AC adsorbent could be potentially applied in a wide range of sample preparation procedures including their miniaturized versions. Other parameters that were not significant in the screening process were kept constant at 10 min, 5 min and 6.5 for ET, DT, and pH, respectively.

Fig. 8
figure 8

Desirability profile indicating the predicted values and desirability functions for neonicotinoids (TCL). MA mass of adsorbent, EV eluent volume

Full size image

Method validation

Under optimum conditions, the analytical performance of the developed d-MSPME procedure was assessed in terms of limits of detection (LOD), quantification (LOQ), linearity, repeatability and reproducibility (Table 3). The neonicotinoids standard solutions ranging from concentrations of 0–700 µg L−1 were preconcentrated using the d-MSPME and the calibration curve for each analyte was constructed. From the calibration curves, the linearity was from 1.4 to 700 µg L−1 with the correlation coefficients ranging from 0.9941 to 0.9987, respectively, which indicates good linearity. The LOD and LOQ were demonstrated based on the International Union of Pure and Applied Chemistry (IUPAC) as the lowest amounts of the analytes that can give responses equivalent to 3.0 or 10 times the original signals, respectively. They are expressed as the ratio of the standard deviation to the slope of the calibration curve (Mpupa et al. 2019; Komendova 2020). The LODs ranged from 0.41 to 0.82 µg L−1, while LOQs were between 1.4 and 2.7 µg L−1. The repeatability (intra-day) and reproducibility (inter-day) procedures were conducted to assess the precision of the method. These were expressed in terms of relative standard deviations (%RSD) based on ten replicates on the same day (repeatability, n = 10) and five different days (reproducibility, n = 5 working days). The intra-day precisions were found to be less than 1%, while the inter-day precisions were less than 6% (Table 3).

Table 3 Analytical performance of the developed method
Full size table

Adsorbent reusability

The attractive feature of an exceptional adsorbent includes high regeneration and reusability. This allows the reduction in the overall cost of the adsorption technology. Additionally, for economic reasons, the desorption of the analyte from an adsorbent is considered as significant approach to the reuse of spent adsorbents. In this study, the reusability studies were carried to determine the number of cycles a spent Fe3O4@PABA/AC nanocomposite can be used for the preconcentration of the neonicotinoids while maintaining maximum extraction recoveries. As seen in Additional file 1, Fig. S2, quantitative recoveries (≥ 80%) for all neonicotinoids were obtained when the adsorbent was used for at least 10 times with the %RSD ranging from 2.5 to 3.1%. Even though the adsorption efficiency decrease was observed, the spent adsorbent showed better stability and reusability. Therefore, it can be concluded that Fe3O4@PABA/AC nanocomposite had acceptable regeneration and reusability performances and can be viewed as a promising adsorbent for efficient extraction and preconcentration of various organic pollutants in water systems.

Application to real samples

The reliability of the d-MSPME method was conducted using the spike recovery tests. These tests were carried at two levels, 1.0 µg L−1 (lower level) and 5.0 µg L−1 (higher level) for all analytes (Table 4). It was clearly observed that the analytes of interest were not detected in all matrices (influent, effluent, and river water samples). The spiking process was done due to the absence of the certified reference material (CRM) for neonicotinoids. Each sample was analysed three times, and the response was obtained from the average of the outcomes. Maximum percentage recoveries (%R) were still obtained in the ranges of 86.7 and 99.2% for all analytes in all matrices, indicating that the matrices had no disadvantageous effects on extraction of the neonicotinoids. The relative standard deviation was < 5%, further proving that the effects from the matrix were very little for the proposed analytical procedure. The d-MSPME analytical method showed potential in extraction and preconcentration of trace neonicotinoids from various environmental wastewater matrices.

Table 4 Application of the suggested analytical method in real water samples (mean ± SD)
Full size table

Comparison between the current proposed method and others

The performance of the current analytical method coupled with HPLC–DAD towards the preconcentration of neonicotinoids was compared with the methods reported recently. The parameters of the present proposed method and those of the previously reported methods are summarized in Table 5. The current proposed method had a good repeatability, and the relative standard deviation (RSDs) were smaller than those of the recently reported methods. Limits of detection of the presented method were lower than some of the mentioned methods and the linear ranges were reasonably wider. Compared to a relative method reported by Ghiasi et al. (2020) for the preconcentration of neonicotinoids, it can be concluded that the current method is facile, rapid, sensitive, and environmentally friendly towards the extraction and determination of neonicotinoids from various environmental samples. Furthermore, the method was able to provide maximum %R for four analytes in the presence of the interferences, and the %R was maintained even after ten successive cycles (Additional file 1: Fig. S2).

Table 5 Analytical characteristics of the proposed method in comparison with the recently reported methods for the extraction and determination of neonicotinoids in various matrices
Full size table

Conclusions

Magnetic-based PABA/AC nanocomposite was successfully synthesized and applied as an adsorbent in the extraction and preconcentration of the selected neonicotinoids (ACT, IMD, TCL and CLD) from model solutions and real water samples. The d-MSPME showed a great potential as a preconcentration method for the selected neonicotinoid insecticides. Fe3O4@PABA/AC nanocomposite was characterized using FTIR, SEM, TEM, BET and XRD. Before the application of the d-MSPME method in the real samples, parameters such as pH, ET, MA, DT and EV affecting the performance of the method were optimized for the extraction of neonicotinoids from spiked tap water. Under optimum conditions, the reliability of the d-MSPME method was evaluated based on LODs and LOQs, which ranged from 0.41 to 0.82 µg L−1 and 1.4 to 2.7 µg L−1, respectively. Fe3O4@PABA/AC showed high reusability which was confirmed by reusing the adsorbent at least 10 times while maintaining maximum recoveries. Furthermore, the adsorbent was used in the extraction of neonicotinoids from real water samples obtained from domestic WWTP. The adsorbent showed magnificent applicability in the neonicotinoids extraction, and this indicated that the nanocomposite might mitigate the global challenge based on monitoring trace organic pollutants in environmental wastewater.