Simultaneous extraction of six phthalic acid esters from polyethylene terephthalate (PET) bottled water using poly (ionic liquid) functionalised silica coated-iron oxide nanoparticles: a risk assessment study

ABSTRACT In this research, a poly (ionic liquid) functionalised silica-coated magnetic nanoparticles were synthesised for the first time and utilised as a novel adsorbent for the simultaneous extraction of six phthalic acid esters from polyethylene terephthalate bottled drinking water samples. The adsorbent was prepared by grafting poly (1-benzyl-3-vinyl-1 H-imidazol-3-ium chloride) onto silica-coated Fe3O4nanoparticles via Cu(0)-mediated reversible-deactivation radical polymerisation method. Various techniques were used to characterise the synthesised nanomaterials. A qualitative analysis was performed using ultra-performance liquid chromatography triple–quadrupole tandem mass spectrometry. Response surface methodology based on central composite design was used to optimise the extraction procedure and the optimised condition for each factor was; pH of3.9, 22 mg of adsorbent, 3.5 min extraction time, 3 min desorption time, 340 µL of elution solvent (methanol).The results of the validation of the method indicated its acceptable accuracy (88.6-100.1%), good linearity (r  > 0.995), satisfactory repeatabilities (RSDs ≤ 6% for intra- and inter–day precisions) and high enrichment factors (535–572). The limits of detection and limits of quantification of the proposed method achieved were 1.1-4.6 ngL–1and 3.6-15.3 ngL–1, respectively. In this study, six common PAEs, including dimethyl phthalate, diethyl phthalate, dibutyl phthalate, benzyl butyl phthalate, bis (2-ethylhexyl) phthalate, and di-n–octyl phthalate, were found in PET bottled water within the range of 0.21-0.94 μgL–1, under different storage conditions. Nevertheless, only a negligible risk is caused by the PAEs in PET bottled water for consumers following the recommendations, such as storing at a common place (25°C), in a short period away from the sun.


Introduction
Drinking water stored in polyethylene terephthalate (PET) bottles has become famous due to its low cost and convenience.Transparency market research predicted fast growth in the global market of bottled water (6.44% per annum) within 2017-2024 [1].Nevertheless, the potential leakage of chemicals like phthalic acid esters (PAEs) from the PET bottle into the drinking water and their potential endocrine-disruption impacts has raised public health concerns [2,3].The flexibility and softness of plastic products can be improved by the use of PAEs as plasticisers [4].Nevertheless, PAEs are leached from thermoplastic packages into drinking water, beverages, and foods [5].For instance, three PAEs compounds, including di(2-ethylhexyl) phthalate (DEHP), dibutyl phthalate (DBP), and diethyl phthalate (DEP), were found in water in the PET bottles [6].
Furthermore, the concentrations of DEHP, DEP, and DBP in the water content of the PET bottles are almost 20 times greater than the water stored in the glass bottles [7].PAEs leakage from a PET bottle is under the influence of storage circumstances such as time, pH, and temperature [8,9].PAEs contamination of the water stored in the PET bottles can be assigned to first, water pollution in the bottling plant; second, PAEs migration from bottle materials to water, and third cross-pollution over-analytical technique [3].Regarding the incrementing use of PET bottled drinking water worldwide, it is essential to determine the PAEs sources in PET bottled drinking water and assess its potential health risks.Among the PAEs, DEHP attracts the most health and environmental concerns, since it has been classified as a possible carcinogen to humans (i.e.Group 2B) by International Agencyfor Research on Cancer (IARC) [10].The World Health Organization (WHO) has established a guideline limit of 8 μgL -1 forDEHP in drinking water [11].United States Environmental Protection Agency (USEPA) hasdefined the maximum limit of DEHP as 6 μgL -1 in the 'National Primary Drinking Water Regulations' [12].However, water quality criteria for some other PAEs such as BBP have not been established by most countries.
Several methods have been developed to determine PAEs among which chromatography-tandem mass spectrometry (GC-MS/MS), high-performance liquid chromatography ultraviolet detector (HPLC-UV), gas chromatography-mass spectrometry (GC-MS), and high-performance liquid chromatography-mass spectrometry (HPLC-MS) can be mentioned [13].
Sample preparation is one of the most significant actions in the process of analysing the analytes [14].Low concentrations of PAEs make their direct detection difficult; necessitating a pre-concentration stage [15].Various extraction processes have been utilised to analyse phthalates in beverages, such as solid-phase microextraction and solidphase extraction processes, along with their liquid phase microextraction methods [16].Magnetic solid-phase extraction (MSPE) is a procedure-oriented method utilising magnetic sorbents to separate and pre-concentrate different analytes from various quantities of the sample [17].The magnetic sorbent is combined with the sample to absorb the analyte [18].An external magnetic field can be used to separate analyte-magnetic sorbent particles from the sample solution, followed by eluting with the perfect eluent.Recently, MSPE sorbent is focused precisely on magnetic nanoparticles (NPs) with different functional groups for effective isolation and/or enrichment of the analytes, from samples with complex matrices [19].
For preparation of the sorbent, usually iron oxide (Fe 3 O 4 ) NPs as the magnetic core of the sorbent are synthesised and coated with different reagents to prevent the particles oxidation in contact with air and increase the selectivity of the sorbent towards the studied analytes [20].
Ploy (ionic liquids) (PILs) include ionic liquid (IL) types within repeating units [21].The unique properties and designability of ILs enhance the applications and properties of PILs in polymeric materials [22].A versatile and robust synthetic method is required to graft PIL over substrates, metal-mediated reversible-deactivation radical polymerisation (RDRP) [23].PILs with a reasonably narrow molecular weight distribution could be obtained by metal-mediated RDRP.A zero-valent metal can be used as a catalyst to simplify the polymerisation process, considerably facilitating large-scale production [24].
Conventional optimisation methods, namely one-factor-at-a time(OFAT), have an extensive application.OFAT is favoured by non-experts, due to its simplicity, especially in situations where the data is cheap and abundant, but it suffers from some disadvantages viz.requiring a large number of experimental runs and chemical waste production, labour effort and being tedious and high cost.Moreover, these methods are incapable of differentiating between the significance of each variable, and as a result of that, the influences of interaction between variables are ignored, and the correct optimum level cannot be achieved.To overcome these impediments, multivariate optimisation is an appropriate and valid statistical method.Response surface methodology (RSM) is a dominant statistical-based strategy to evaluate the effects of different factors simultaneously which enables the optimum conditions for providing the desirable response [25,26].
In this research, PIL functionalised magnetic silica coated-magnetic Fe 3 O 4 NPs were synthesised.1-benzyl-3-vinyl-1 H-imidazol-3-ium chloride as an ionic liquid monomer was synthesised by the reaction between 1-vinylimidazole and benzyl chloride in the presence of ethanol and 2,6-di-tert-butyl-4-methylphenol as a stabiliser.This monomer was then used for the first time for polymerisation of Fe 3 O 4 /SiO2 NPs' surface using the Cu(0)mediated RDRP technique (Scheme 1).Various techniques were utilised for the characterisation of synthesised nanomaterials.The performance of the poly (1-benzyl-3-vinyl-1 H-imidazol-3-ium chloride) functionalised silica coated-magnetic iron oxide nanoparticles(Fe 3 O 4 @PIL NPs) as a novel sorbent was evaluated in the MSPE of six PAEs from water samples followed by quantitation with UPLC-MS/MS technique.The parameters affecting extraction efficiency were optimised by RSM based on central composite design.The proposed method was used to analyse the PAEs in PET bottled water samples to understand the leaching characteristics of PAEs from PET bottles into the water.The health risk associated with PAEs leakage into the water was explored.To the best of our knowledge, this research is the first study investigating nanoparticles' application in health risk assessment study.

Materials and method
Materials, instrumentation, UPLC-MS/MS conditions, sample preparation, brand selection and consumer Survey,measurement of total PAEs in PET bottles, storage conditions of PET bottled water, the effect of temperature on the concentration of PAEs in bottled water, synthesis of 1-benzyl-3-vinyl-1 H-imidazol-3-ium chloride as a IL monomer ([BVI]Cl), synthesised of Fe 3 O 4 @PIL NPs, extraction method andhealth risk assessment are described in Supplemental online material(Table S1-S3, Figure S1).

Characterisation of NPs
FESEM was utilised to characterise the size and morphological structure of Fe 3 O 4 @PIL NPs. Figure 1(a) represents the spherical NPs with a particle size of about 35 nm. Figure 1(b) represents the TEM image of Fe 3 O 4 @PIL NPs.Here, Fe 3 O 4 NPs, the solid black sphere, provides the magnetic MSPE sorbents with a superior magnetic response.The PIL layer is represented by the light shell protecting the central magnetic core from corrosion and oxidation and endowing robust affinity to the objective analytes.DLS was performed to clarify the particle size distributions of the NPs in the solvents and interactions between the particles and the solvents.For Fe 3 O 4 @PIL NPs,the average hydrodynamic diameter of 140 nm was obtained, and it was 25 nm for Fe 3 O 4 in ethanol utilising DLS.Such results were considerably different from the outcomes of FESEM and TEM attained in dry vacuum conditions.The agglomeration, polarity, and swellingof the particles in the solvents influenced differently the experiment results [27].To characterise the Fe 3 O 4 NPs, Fe 3 O 4 @SiO2 NPs, and Fe 3 O 4 @PIL NPs, the transmission mode FTIR spectrum was recorded in the wavenumber range of 4000-400 cm-1 (Figure 1(c)).According to the FTIR spectrum for the Fe 3 O 4 NPs, there is broadband at 3475 cm−1 due to the stretching vibration of surface -OH groups, and the weak band at 1616 cm−1 is attributed to the hydroxyl bending and/or adsorbed water on the sample surface [28].The Fe-O stretching vibration resulted in  C-H stretching vibrations appeared at 2930 cm-1, and 2860 cm-1, respectively, and the bending vibration absorption peaks of -CH2 and -CH3 appeared in 1462 cm-1 and 1356 cm-1, respectively [32].Considerably, the novel band around 1430 cm-1 is allocated to the C-N +group stretching vibration within magnetic NPs modified by ILs, which indicates the effective ILs-modifying the Fe 3 O 4 @SiO2 NPs surface [33].

Optimisation of MSPE parameters
The optimisation experiments were performed in triplicate, and 50.0 mL of spiked UPW was used that contained a concentration of 5.0 μgL -1 of each target analyte.Extraction recovery (ER) was utilised to investigate the optimum condition [35].The percentage of the total analyte extracted into the sedimented phase defines ER.As an analytical response, the ER was calculated using Equation (1): where C0 and Csed are the initial concentration of the analyte in the sample and the concentrations of analyte in the sedimented phase, respectively.C0 has been determined bases on a calibration curve obtained utilising direct injection of standard solutions.Vsed and Vsam are the volumes of the sedimented phase and aqueous sample, respectively.Optimisation of the elution solvent is described in Supplemental online material (Figure S2).

Optimisation by central composite design
Three groups of experimental designs, such as full factorial, 1/2 fraction factorial, and small, are included in a Box-Wilson central composite design (CCD) (See Supplemental online material) [36].It suggests an appropriate curvature surface model while considering parameter interactions.To guarantee a minimum number of runs and high accuracy, this research selected a 1/2 fraction factorialdesign that includes five factors at five levels: sample pH (A), sorbent amount (B), contact time (C), desorption time (D), and elution amount (E) (Table S5).The range of each factor was chosen based on single-factor experiments, which were designed, analysed, and optimised using the Design-Expert® Software Version 11 (Stat-Ease Inc., Minneapolis, MN, USA).According to Table S6, in total, 32 runs exist (which can be determined from N = 1/2 2k + 2k + N0, where k and N0 denote the number of central points and parameters, respectively).The central points were then used to determine the data's reproducibility and the experimental error.In the present study, the imitation scores and the responses (Yi; ER %) were estimated for every required experimental trial based on the applied mathematical modelling.Least squares regression was also used for fitting the results to a second-order polynomial equation, as represented in Table S7.The quality of the fitted model can best be evaluated by analysis of variance (ANOVA), a statistical technique used to analyse experimental data and determines the proportion of influence of a factor or a set of factors on total variation.The statistical significance of the aforementioned partial quadratic model was evaluated based on the F-test and ANOVA results demonstrated in Table S8.The ANOVA results of DMP (as an example) are discussed in the following.A similar result was obtained for other analytes.
The regression model's high statistical significance is best manifested in F value = 133.63(Fisher's F-test value) and 0.0001 (p-value).This demonstrates that the five variables above and the experimental data obtained from extraction recovery are correlated.Furthermore, the quality of the fit quadratic polynomial model and its overall predictive capability is evaluated by coefficient of determination (R 2 ), adjusted R 2 (Adj.R 2 ), and CV [37].An R 2 of 0.9959 means that the five independent variables above account for 99.59% variations in the extraction recovery, which can be justified by a model with a CV of 2.21%.Accordingly, the model could explain 99.59% of total variations.As can be seen from Figure S3, an adjusted R 2 of 0.9884 (high to advocate a high significance of the model) indicates independent variable alone is responsible for the explanation of 98.84% of data points in the regression line estimated, influencing the dependent variable practically, yet adjusted for the number of model terms.There are a high correlation and logical agreement between an adjusted R 2 of 0.9884 and a predicted R 2 of 0.9157.This indicates that the predicted and experimental values for goodness-of-fit of the regression equation and ER are highly consistent.The model has a lack-of-fit of 2.21 (significant, p-value < 0.05), which suggests that the difference between actual values and predicted values is influential in the pure error between the replicates [38].
Response surface analysis(taken DMP as an example) are described in Supplemental online material (Figure S4).
The numerical optimisation of the statistical software was selected to find the specific point that maximised the global desirability function (DF) (see Supplemental online material).According to the response optimiser data, the CCD optimisation design matrix indicated that the responses were maximum at these conditions: pH of 3.9, 22 mg of adsorbent, 3.5 min extraction time, 3.0 min time of desorption, 340 µL of elution solvent.The DF was 1.0, which represents an entirely desirable or ideal response value of these conditions [39].Under the optimised conditions, the predicted ER percentage values were varied from 93.4% to 99.8%.The verification experiments were carried out under similar circumstances.The corresponding actual experimental values of ER% were determined to be in the range of 91.9-100.2%(n = 5), which is closely related to the data of the predicted values, proving the reliability of RSM optimisation.

Sorbent reusability
An eco-friendly economical procedure is ensured mainly by reusability, according to which adsorption performance is assessed.Several successive cycles of adsorption/desorption were performed during each run, of which ≈95% sorbent was recovered effectively.Once experimenting four cycles, about 85% of ER values were obtained, indicating the applicability of Fe 3 O 4 @PIL NPs without a substantially lost extraction efficiency.In the fifth run, however, ERs considerably declined, i.e. 49%.The NPs mass declined upon some regenerations due to being rinsed during recycling and reuse, and modified polymercoated Fe 3 O 4 NPs lost their efficiency upon a couple of rising attempts, leading to limited NPs reusability.Accordingly, the reusability of Fe 3 O 4 @PIL NPs was limited to five cycles, potentially resulting in significantly dropped extraction efficiency.

Method validation
The linearity, sensitivity, precision, accuracy, carryover effect and enrichment factor (EF) of the established method were investigatedunder optimum conditions.

Linearity range and sensitivity
Calibration curves were created by at least six calibration concentrations of every analyte, in which y and x represent the response ratio of analytes to IS and the concentrations of analytes, respectively.Sensitivity was assessed using the limit of detection (LOD) and limit of quantification (LOQ).LOD and LOQ were obtained by the signal-to-noise response ratios (S/N) of 3:1 and 10:1 in respective order [40].Table 1 summarises the calibration curves, correlation coefficients, linear ranges, LOD, and LOQ of the analytes.The whole compounds are well related linearly in the limit of the chosen linearity with correlation coefficients above 0.995.

Accuracy and precision
The closenessbetween repetitive individual measurements of QC samples is described using precision, represented as the coefficient of variation (CV), and the proximitybetween measured and actual values is described using accuracy, represented in the recovery.Considering the lack of certified reference materials, these investigations were carried out for QC water samples.Six repetitious QC samples on the same day were measured, and the CV and recovery values were calculated to evaluate intra-day precision and accuracy, which were measured by assessing six repetitious QC specimens on three successive days and estimating CV and RE values.Intra-day and inter-day precisions and accuracy need to be assessed at four concentrations (LOQ, LQC, MQC, and HQC).
Table S9 summarises the intra-day and inter-day precision and accuracy at the four concentrations.The intra-day and inter-day precisions (CV) of the entire analytes ranged from 2.01% to 5.77%, and the accuracies (recovery) ranged from 88.6% to 100.1%.The above observations exhibited the accuracy and reproducibility of the demonstrated method.For instance, Figure 2 reveals the chromatogram of compounds within the studying sample spiked with 2 µgL -1 for each PAEs using the developed method.Also, characteristic chromatograms of blank sample are depicted in Figure S5.

Carryover effect
Immediate injection of the blank sample was done following the upper limit of quantification (ULOQ) concentration of the calibration curve.A water sample carried out injections with a high concentration of standard curve, and then by the blank sample.The carryovers in the blank water samples were less than 0.08% and 0.06% of the analyte responses at the ULOQ, and for IS, respectively, suggesting no considerable impacts on the results.

Enrichment factor
Increasing sample volume can improve the enrichment factor, but high volume may lead to limited dispersion.The impact of sample volumes was examined under fixed mass of the analytes and various feed volumes within the range of 50-500 mL.The extraction recoveries of analytes were quantitative up to 200 mL of the sample volume.Moreover, in MSPE, EF was defined as the ratio of analyte concentration in the sedimented phase after extraction to the initial analyte concentration in the aqueous phase, which was used to evaluate the enrichment efficiency of MSPE.EF value for each analytewas shown in Table 1.

PAEs' concentrations in commercial bottled water
All six PAEs were found in the ten various commercial bottled water brands stored in outdoor circumstances (Figure 3(a)).The PAEs overall concentrations were within the range of 0.235-0.937μgL -1 with a mean value of 0.425 μgL-1.Correspondingly, the PAEs' concentrations in PET bottled water stored under indoor circumstances were in the range of 0.202-0.626μgL -1 , with a mean value of 0.346 μgL -1 (Figure 3(b)).Within the ten brands, the maximum PAEs' concentrations were found in brand B4, regardless of their storage in outdoor or indoor conditions.There was no significant difference between PAEs concentrations in water stored outdoors or indoors (p > 0.05).Nevertheless, the PAEs' total concentration in brands B4 and B5 represented a weak incrementing trend, mostly associated with increasing DBP concentration.Within such compounds, DMP and DBP were the most abundant and common PAEs, with over 60% of the total PAEs determined in all the specimens.In comparison to the accessible data, the DEHP's concentrations were lower compared to WHO standards of 8 μgL -1 [11] and USEPA standards of 6 μg.L-1 [12].The overall PAEs' concentrations in commercial bottled water stored in PET bottles at 50, 60 and 70°C were, 0.287 ± 0.038, 0.425 ± 0.087 and 0.516 ± 0.088 μgL -1 , respectively (Figure 3(c)).Furthermore, a weak negative association was found between the quantity of PAEs and bottle thickness in commercial water (Figure S6).Nevertheless, the PAE concentrations are still unaltered by increasing the temperature, and the PAE concentration was not dependent on water types.There was no significant difference amongst compounds in water stored in PET bottles from the ten brands, excluding brands B8 and B5 representing a weak incrementing trend (p-value of higher than 0.05).Correspondingly, the concentration of DMP, DEP, and DBPwith more than sixty percent of overall PAEs in all the specimens, among which the DBP was the most plentiful compound, and the DEHP concentrations were lower compared to the accessible WHO standards (8 μgL -1 ), and USEPA (6 μgL -1 ) Figure 3. (a) Total concentrations (ngL -1 ) of PAEs in each brand of PET bottled water stored outdoors for 6 weeks (n = 3), (b) Total concentrations (ngL -1 ) of PAEs in each brand of PET bottled water stored in laboratory for 6 weeks (n = 3), (c) Total concentrations (ngL -1 ) of PAEs in each brand of PET bottled water after incubation at 50, 60 and 70°C for 24 h (n = 3).[11,12].Concentrations of PAEs in pure water stored at different temperatures is described in Supplemental online material (Figure S7).DBP, DEHP, and DEP leakage from thermoplastic products was investigated owing to their potent consequences on the endocrine systems [3,41].Former evidence indicated the release of DEHP, DBP, and DEP from PET bottles into water under different circumstances (long storage times, sunlight, and high temperature).Likewise, these results also indicated the leakage of DBPs into the water from PET bottles kept at higher storage temperatures and time, though no positive relationships were found between the storage circumstances and PAEs concentrations.Furthermore, DBP mainly contributed to increasing the PAEs concentrations.Such findings were in line with the results of Diana and Dimitra [3] and Keresztes et al. [42].The PAEs concentrations in water stored in PET bottles were not only affected by temperature.The results reveal other PAEs sources in addition to the PET bottles themselves.Diana and Dimitra [3] stated that several pathways might lead to the presence of PAEs in bottled water, including the first leakage from bottle substances with low quality, contamination in bottling in plants due to use of plastic pipes, and experimental pollution over the analytical process owing to the extensive utilisation of plastic instruments.
The current work investigated the PAEs source in water specimens by assessing PAEs in the PET bottles plastic material and the contained water. 2 or 3 types of PAEs were found in the 10 PET bottles; however, six PAEs were observed in water stored in the same brand bottles.These results suggest that PET bottles are not the only source of PAEs in the bottled water.Thus, the water might be contaminated with PAEs before bottling.Therefore, more precise measurements should be considered to remove the potential risks of chemical compounds within PET bottled drinking water over the entire production procedure [43].
Furthermore, three kinds of PAEs were found in ultra-pure laboratory water whose concentrations were incremented by the storage time.More PAEs were found as well.Such findings indicate the transfer of some PAEs from PET bottles to water.Nevertheless, leakage seemed to be not dependent on storing temperature within a short period.Thus, PAEs found in water stored in PET bottles were more probably affected, at least partially, from the bottles themselves.Prominently, the DEHP concentrations were lower than USEPA (6.0 μgL -1 ) and WHO (8.0 μgL -1 ) standards, even after 24 hours of storage at 70°C.However, Ceretti et al. [44] indicated no trace of DBP or DEHP in water stored in PET bottles, even after storage for 10 days at 40°C.This inconsistency might be related to the differences in bottle composition.Ultraviolet (UV) light may be another reason in this regard [3,6].According to the previous study, the PAE (DEP, DEHP, and DBP) concentrations were increased under the poor storage circumstances in PET bottles, possibly due to the incremented UVexposure from sunlight [7].Furthermore, UV radiation enhanced the leakage of different chemicals from plastic bottles and other substances like PAEs and additives [45][46][47].Both outdoor and indoor simulation examinations were conducted with no light to retard the PAEs release.The recommended storage circumstances to preserve the quality of bottled water involve keeping the bottled water at a commonplace (25°C) for a short time away from the sunlight.

Health risk assessment
The adverse health result risk was assessed in terms of four types of PAEs (DEP, DBP, BBP, andDEHP) in PET bottled water with accessible RfDs.For such PAEs, HQs related to the use of water stored in PET bottles were < 1 (Table 2).As estimated, most customers would encounter insignificant health risks from such four PAEs in PET bottled water by drinking two bottles of water each day stored under normal circumstances.
The findings indicated that the health risk related to the levels of selected PAEs was acceptable for adults despite the PAEs source in PET bottled water.This was consistent with a former report [9].Nevertheless, this assessment should be cautiously interpreted since humans are subjected to PAEs via multiple ways along with bottled drinking water.It is indicated that humans can be subjected to further PAEs, and such mixtures may partially act additively to produce toxicity.Thus, it is essential to evaluate the PAEs' potential health risks in future works considering the continuously increasing use of plastic-bottled water [45].

Comparison of proposed MSPE method with previous methods
In comparison to the approaches that were previously reported, the current procedure as available concerning linearity, precision, LOD, used solvent, and amount of sorbent.A brief assessment is represented in Table 3 for comparing the current method with those previously employed in the extraction of PAEs from water samples.The established method represented the benefits of extensive linearity range, desirable accuracy, and extreme sensitivity.Furthermore, it should be noted that the developed method contained lower LODs in comparison to the other previously reported approaches revealing exceptional extraction efficacy of the proposed MSPE method.In comparison to the investigations, the developed method included the RSD values marginally or considerably lower.In this work, only small quantities of core-shell Fe 3 O 4 @PIL NPs were required for operative extracting of PAEs, possibly owing to their superior adsorption capability.Furthermore, the adsorbents can be reutilised for several times.A limited organic solvent was only used (without water-immiscible dangerous solvents), making it an eco-friendly method.Such results reveal that the proposed high efficiency and sensitivity can determine PAEs at the trace levels within complex matrices.

Conclusion
The current study represented a versatile and effective adsorbent containing imidazolium-based PIL-functionalised Fe 3 O 4 NPs.Furthermore, it was utilised as an MSPE to extract and preconcentrate PAEs from PET bottled drinking water prior to their analysis via UPLC-MS/MS.Validation experiment outcomes found the satisfactory precisions within the linear ranges.The synthesised nano-sorbent is an innovative and attractive candidate for substituting the traditional MSPE adsorbents to extract these compounds from the complex matrices.Hence, lower amounts of adsorbent are used for obtaining reasonable extraction effectiveness.Thus, utilising less volume of organic solvents that is highly favourable within Green Chemistry becomes possible.In water in PET bottles stored under common storage, PAEs including DMP, DEP, DBP, BBP, DEHP, and DNOP, ). were detected using developed methods.The investigation showed that the plastic bottles were not the sole source of PAEs, and stricter measurements should be taken to reduce the potential health risk of PAEs in PET bottled drinking water.

Figure 1 (
d) shows the XRD pattern of Fe 3 O 4 NPs, Fe 3 O 4 @SiO2 NPs, and Fe 3 O 4 @PIL NPs, along with the diffraction peaks at 2θ = 35.3º,41.5°, 50.5°, 63.4°, 67.5°, and 74.5° assigned to (220), (311), (400), (422), (511) and (440) planes, respectively.It indicates a reverse cubic spinel phase of Fe 3 O 4 NPs based on the standards in magnetite structure studies (JCPDS card NO. 85-1436)[34].The amorphous layers found within the range of 15 to 25° prove the presence of a silica layer over the Fe 3 O 4 @SiO2 NPs surface.Considering the XRD pattern of Fe 3 O 4 @PIL NPs, the prominent peak of modified NPs is correspondent with the peaks of bare Fe 3 O 4 NPs, which indicates that grafting polymerisation created no change in the crystalline structure of Fe 3 O 4 NPs.The NPs' magnetic properties were evaluated utilising the VSM at room temperature.There was no hysteresis in Figure1(e), revealing the NPs' superparamagnetic performance at RT.That is, no magnetism is kept in the nanomaterials following eliciting the magnetic field.The pure Fe 3 O 4 NPs magnetising saturation of 67 emug-1 was obtained and reduced to 34 emug-1 followed by the polymer coating process.Polymerisation had no impact on the magnetic properties of the super magnetic Fe 3 O 4 @PIL NPs because the Fe 3 O 4 NPs structure was not changed in this process.Measuring zeta potential at different pH values, the pH of the NPs' isoelectric point was determined (Figure1(f)).The Fe 3 O 4 @PIL NPs' zeta potential indicated a positive charge for most values ofpH.This surface with a positive charge is caused by quaternary ammonium moieties resulting in creating positive types at most pHvalues.The BET technique was used to measure the specific surface areas of the NPs by measuring the amount of the adsorbed N2 gas on the NPs' surface.The N2 adsorption/ desorption isotherms of the Fe 3 O 4 @PIL NPs that exhibit type IV indicating the mesoporous characters of the synthesised polymer-modified-NPs.

Table 1 .
Calibration data, LOQ and LOD of developed method.