Modication of Magnetic Graphene Oxide By An Earth-Friendly Deep Eutectic Solvent To Preconcentrate Ultratrace Amounts Pb(II) In Oil Seeds

The aim of this article is presenting an earth-friendly deep eutectic solvent (DES) to preconcentrate ultratrace amounts of Pb(II) prior to its quantication by ame atomic adsorption spectroscopy. The synthesis of adsorbent started by preparing graphene oxide according to the modied Hammer’s method, followed by magnetization by Fe 3 O 4 nanohemispheres. Magnetic graphene oxide was dispersed in a mixture of LiCl and urea at 60° via ultrasonication. All the materials are environmentally-friendly and the preparation strategy is energy ecient. X-ray diffraction, scanning electron microscopy, alternating gradient force magnetometer and Fourier-transform infrared spectroscopy were applied to characterize the products. Graphene oxide has a large surface area and could be functionalized with DESs through π-π interaction and electrostatic force. Urea has active negative sites, which garb heavy metals due to the interaction between negative and positive agents. Accordingly, this adsorbent (UreaLiCl-mGO) could be offered as a capable adsorbent to preconcentrate ultratrace amounts of Pb(II). Conditions were optimized, and under the optimum situation, (a) limit of detection of 99 × 10 -8 g L -1 , (b) relative standard deviation (n=5) of 1.3%, (c) preconcentration factor of 100 (d) linearity of dynamic range of 5.0 × 10 -6 – 23 × 10 -6 g L -1 , (e) durability of 6 months and (f) reusability of 7 times prove applicability of the adsorbent. The tests of selectivity, effect of interference ions, swelling property, isotherm of adsorption, kinetic of adsorption and thermodynamic of adsorption were completely investigated. Four different oil seeds were successfully applied as real samples.


Introduction
Lead is a hazardous heavy metal, which reasons many health problems such as poisoning metabolic, inhibiting enzymes, damaging nervous connections, being replaced with calcium in bones and causing blood and brain disorders [1,2]. Therefore, presenting applicable analytical methods to quantify its ultratrace concentrations are vital [3]. Magnetic solid phase extraction (MSPE) -a novel classi cation of solid phase extraction -is an authentic analytical technique. In this method, analyte was adsorbed onto a magnetic adsorbent, so ltration and centrifuge are substituted by applying an external magnetic eld [4][5][6][7][8].
Using nanomaterials is an actual policy to advance the pro ciency of MSPE [9]. Magnetic nanostructures have stronger magnetic properties, leading to simplify separation steps and decreasing separation time. Moreover, nanoadsorbents have more surface area in comparison with non-nanomaterials; since, these adsorbents have more a nity toward analyte(s), as well as being functionalized by other agents [10][11][12].
Deep eutectic solvent (DES) is a new classi cation of ionic liquid groups, but has lower lattice energy and lower melting points. Researchers suggest many types of DESs, but one of them is mixtures of metal halides and urea with melting points of <80. These materials are safe, inexpensive, accessible and green. Therefore, the nal product is environmentally-friendly and the method of preparation is energy e cient [13][14][15][16].
Up to now, a variety of DES have been applied to preconcentrate contaminants. A dispersive liquid-liquid microextraction based on DES, including choline chloride and three different hydrogen bond donor ethylene glycol, oxalic acid or urea was applied to preconcentrate Pb(II) and Cd(II) form various water samples [17]. In another example, methadone, Cd(II), Pd(II), Ni(II) and Cu(II) in biological samples were detected via liquid-liquid extraction method based on DESs [18,19]. In addition, Shokuhi Rad et al. proposed a mixture of choline chloride and 4-boromo phenol as a DES to preconcentrate Ni(II). Gul Kazi et al. preconcentrated Pb(II) and Cd(II) from cosmetic sample based on mixing ZnCl 2 and CH 3 CONH 2 [20].
In this article, a DES of LiCl and urea was loaded onto magnetic graphene oxide (mGO) and then the adsorbent (UreaLiCl-mGO) was applied as a magnetic adsorbent to quantify ultratrace amounts of Pb(II) in oil seeds samples. Synthesis procedure is very facile, fast, low-temperature, low-pressure and environment-friendly. X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscopy and alternating gradient force magnetometer check the crystallinity, loading DES, morphology and magnetic property, respectively. The preconcentration conditions were optimized and analytical gures of merits were reported. A ame atomic adsorption spectrometer measured the concentrations of analyte. Isotherm, adsorption kinetic and adsorption thermodynamics showed the adsorption mechanism of Pb(II) onto UreaLiCl-mGO.

Reagent and solution
All the reagents were of analytical grade. The salts for preparing Pb(II) standard (1.0 × 10 −3 g L −1 ) and testing interference ions, NaNO 3 , graphite powder, KMnO 4 OH, urea and LiCl were procured from Merck Company (www.merck.de, Darmstadt, Germany). Ultrahigh purity from a Milli-Q system water was used to prepare sample solutions.

Synthesis of UreaLiCl-mGO
Firstly, GO was synthesized according to the modi ed Hummer's procedure. Graphite powder (1.0 g), NaNO 3 (1.0 g) and H 2 SO 4 (23 mL) were mixed together in an ice bath and stirred 15 minutes, followed by adding KMnO 4 (3.0 g) at 20°C and stirring for two hours. The mixture was kept at 35°C for one hour and then deionized water (45 mL) was slowly added to the mixture. The suspension was re-stirred at 98°C for 30 minutes. Afterwards the temperature was xed at 25°C to add deionized water (140 mL) and H 2 O 2 (30%, 12 mL) and then the materials were stirred for 2 hours. The product was ltered, eluted with water, eluted with HCl (5%) and dried at 60°C [21].
Finally, mGO was modi ed by DES. A mixture of Urea (5.0 × 10 −2 g) and LiCl (5.0 × 10 −2 g) was heated in a sand bath at 60°C. After obtaining a colorless liquid, mGO (5.0 × 10 −2 g) was dispersed into the DES via one hour sonication. The nal product (UreaLiCl-mGO) was washed with distilled water and then cooled to room temperature [13].

Sample preparation
Sesame, hemp, sun ower seeds and grapeseed were bought from local supermarkets in Tehran. They were cleaned with double-distilled water and dried at 50°C. The cooled samples (1.0 × 10 −1 g) were immersed in HNO 3 (15 mL, 65%) for 48 hours. The samples were heated (90°C), followed by adding H 2 O 2 (20 mL, 30%). The samples were re-heated (150°C) for 10 minutes until ceasing the evolution of fumes. This procedure was repeated until obtaining a clear transparent solution. The samples were cooled to room temperature, ltered with sieve paper and lled to the mark in a 200 mL falcon with distilled water. pH of four separate 50 mL aliquots of digested samples was adjusted in 5 ± 0.5 by HNO 3 and NH 3 solutions. The general procedure was applied as mentioned in section of general procedure. Matrix spiking with standard of Pb(II) (1.0× 10 −5 , 5.0 × 10 −5 and 15 × 10 −5 g L −1 ) was applied to assess the matrix effects.

Characterization
FT-IR analyses con rm that DES was successfully loaded onto mGO (Fig. 1). The FT-IR spectrum of mGO was shown in the black spectrum. The stretching frequencies of the O-H band is illustrated at 3121 cm −1 [22]. The absorbance at 1586 cm −1 is attributed to aromatic C=C. Stretching vibration of C-O groups caused duple bonds at 1117 and 1042 cm −1 [22]. Peaks lower than 700 characterizes bond of Fe-O [23].
The green spectrum shows the FT-IR spectrum of UreaLiCl-mGO. The bands of 3400 cm −1 , 1600 cm −1 and1400 cm −1 are associated to stretching frequency of N-H, deformation frequency of N-H and stretching frequency of N-C, respectively. The sorption peaks of Li-O and Fe-O were seemed at the wavelengths of 1300 cm −1 and 700 cm −1 , respectively [23].
Plot of identi ed phase mGO is shown in Fig. S1. The X-ray diffraction (XRD) pattern of UreaLiCl-mGO is plotted in Fig. 2a the process of modi cation has no consequence on mGO purity and structure. The broad peaks are related to the presence of UreaLiCl. According to the Scherrer formula (Equation 1), the crystal size of the total product decreased from 17 nm to 6 nm, due to applying ultrasonic waves in the process of preparation [13].
(Eq. 1) In this formula, λ is wavelength of incident beam (1.5406 Å), β is FWHM of peak in radian, θ is diffraction angle and k is the Scherrer constant (0.9).
3.2 Optimization of the method 3.2.1 Effect of pH pH is one of the effective parameters on adsorption process. In alkaline pH, heavy metals form oxo or hydroxo complexes with OH − and in acidic pH, H + ions occupy the active negative sites of adsorbents. In both condition the a nity of adsorbents to adsorb heavy metals decreases. To investigate the effect of pH, pH of a series of samples was adjusted from 2 to 9 and the procedure was applied according to section 2.4. The results are plotted in Fig. 3. From pH=4 to pH=6, the recovery stayed maximum. Before pH=4, H + occupied the active sites of adsorbent and after pH=6, Pb(II) ions were precipitated as Pb(OH) 2 and PbO, so recovery decreased. All subsequent works for separation and preconcentration of Pb(II) were done in pH of 5 ± 0.5.

Effect of equilibrium time and amount of adsorbent
The effect of contact time on Pb(II) adsorption onto UreaLiCl-mGO was investigated at different intervals in the range of one to 30 minutes. After gathering UreaLiCl-mGO by a magnet, the supernatant was removed and the settlement was eluted by HNO 3 (1 mL, 2 mol L −1 ). The recoveries are steady after 15 minute so it was selected as the optimum agitation time for all the subsequent experiments.
A series of UreaLiCl-mGO ranging from 1.0 × 10 −2 g to 5.0 × 10 −2 g was tested and 1.0 × 10 −2 g provided satisfying recovery in Pb(II) preconcentration. Low amount of the adsorbent con rms high a nity of UreaLiCl-mGO toward Pb(II) adsorption.

Selecting the best eluent and time of desorption
Elution step was optimized by investigating type and volume of eluents (Table 1). By increasing HNO 3 concentration, the adsorption e ciency increased and HNO 3 of 2 mol.L −1 was selected as the best eluent.
Adsorbent collapsed in the presence of HCl, which is related to its negative consequence on Fe 3 O 4 and DES. Meantime, 0.5 mL, 1 mL, 1.5 mL and 2 mL of HNO 3 were tested and 0.5 mL was optimized volume of eluent.   Effect of desorption time was tested in the range of 1 -5 minutes in room temperature. After 3 minutes, the recovery was maximum, so 4 minutes was selected as the optimum desorption time.

Reusability and durability of UreaLiCl-mGO
In order to check the reusability of UreaLiCl-mGO, several consecutive preconcentration cycles of Pb(II) were examined according to general procedure. No signi cant variation was recorded up to seven cycles; however, preconcentration recovery decreased after this cycle. This weakness is related to the negative effect of HNO 3 onto the adsorbent. H + ions occupy the sites of adsorbent, which decrease the tendency of the adsorbent toward Pb(II).
Oxidation of magnetic adsorbents is one limit in MSPE method, which con nes the applicability of magnetic adsorbents to a short time after their preparation. This feature is named durability and could be achieved by repeating the preconcentration procedure during several days. The performance of the adsorbent was checked during 6 months and the recovery has no important changes. During these months, the adsorbent was kept at room condition and preconcentration procedure was conducted as mentioned, which con rms stability of UreaLiCl-mGO against oxidation. This resistibility is related to presence of DES, which prevents Fe 3 O 4 nanoparticles from collapse.

Effect of interference ions and selectivity of UreaLiCl-mGO
Pb(II) could be adsorbed onto UreaLiCl-mGO via electrostatic interaction, which is not a selective force and may be affected in the presence of other metals [23]. In addition, some onions and Pb(II) may form complexes, and decline the preconcentration ability of UreaLiCl-mGO. This feature was tested in the presence of various concentrations of different cations and anions in test solutions of Pb(II) (50 mL, 5.0 × 10 −5 g L −1 ). The signals were compared with that of a solution containing only Pb(II) and recoveries were in the range of 95 -105%. In concentration levels higher than those usually present in real samples, the method was not impaired ( Table 2). Selectivity was checked via comparing the performances of UreaLiCl-mGO to adsorb Pb(II) (5.0 × 10 −5 g L −1 ) and other heavy metals (5.0 × 10 −5 g L −1 ) at the same time ( Table 3). The ground reason for selective adsorption of Pb(II) onto UreaLiCl-mGO is their strong interaction in comparison with interactions of other heavy metals. Due to the small ionic radius of Pb(II), this ion has high charge density and is a hard ion; accordingly, it occupies the negative sites of adsorbent stronger and faster.

Sample analysis
Four different seed oils including sesame, hemp, sun ower seeds and grapeseed were tested as real samples. The preparation method of samples was expressed in section of sample preparation. The concentrations of the analyte were determined in both the spiked and unspiked samples according to the following equation.
(Eq. 3) Where, C 1 , C 2 , C 3 and R% are spiked portion, unspiked portion, the concentration of the analyte and relative recovery, respectively. According to data, the method is highly applicable to preconcentrate Pb(II) in seed oil (Table 4).
UreaLiCl-mGO is compared with some new adsorbents (Table 5). UreaLiCl-mGO has signi cant improvements in adsorbent amount, RSD and LDR. Additionally, UreaLiCl-mGO is applicable in semineutral pH, which decreases the usage amount of solutions to pH adjustment. More importantly, UreaLiCl-mGO is a green, safe and earth-friendly adsorbent.

Adsorption isotherm
Adsorption isotherm investigates the amounts of analyte adsorbed per unit mass of adsorbent and the nature adsorption of analyte onto adsorbent [31]. According to Fig. S3a, by increasing the initial concentration of adsorbate, the amount of Pb(II) taken by unit mass of UreaLiCl-mGO increased. Four different isotherm models including Langmuir (Fig. S3b), Freundlich (Fig. S3c), Temkin (Fig. S3d) and Dubinine-Radushkevich (Fig. S3e) were considered to investigate the adsorption model [31,32]. Freundlich, a model of multilayer adsorption, is tted with the system (Table S1). Respective data are given in Electronic Supplementary Materials.

Adsorption kinetic
The relation between time and adsorption of Pb(II) onto UreaLiCl-mGO is investigated through kinetic of adsorption. Two models of Pseudo-First-Order (Fig. S4a) and Pseudo-Second-Order (Fig. S4b) were studied (Table S2). The Pseudo-First-Order model is tted with the system. The pseudo-rst-order kinetics equation describes the adsorption in solid-liquid systems based on the sorption capacity of solids. It assumes that one ion is adsorbed onto one unoccupied adsorption site on the UreaLiCl-mGO [33,34]. Respective data are given in Electronic Supplementary Materials (ESM).

Adsorption thermodynamic
The relation between temperature and adsorption of Pb(II) onto UreaLiCl-mGO is studied through thermodynamic of adsorption. The temperature has positive effect onto adsorption capacity of UreaLiCl-mGO (Fig. S5a). The positive values of ΔH and ΔS show endothermic nature of adsorption and more agitation in sample solution. Negative values of ΔG con rm spontaneous adsorption of Pb(II) onto UreaLiCl-mGO [35,36]. The calculated data are achieved according to Fig. S5b. They are summarized in Table S3. Respective data are given in Electronic Supplementary Materials (ESM).

Conclusion
This article investigated the applicability of a DES of Urea and LiCl, loaded onto mGO to preconcentrate trace amounts of Pb(II). The preparation method is safe, quick, green and facile and the reagents are inexpensive and accessible; accordingly, the method is environmentally-friendly and energy e cient. The characterizations were conducted via FT-IR, XRD, SEM and AGFM. The swelling property was investigated. The optimum condition is pH of 5±0.5, adsorbent dose of 1.0 × 10 −2 g, adsorption time of 15 minutes, eluent of HNO 3 and desorption time of 4 minutes. Pb(II) was quanti ed by a FAAS. LOD (99 × 10 −8 g L −1 ), LOQ (33 × 10 −7 g L −1 ), RSD (1.3%), PF (100) and LDR (5 × 10 −6 -23 × 10 −5 g L −1 ) con rm practical applicability of UreaLiCl-mGO to Pb(II) preconcentration. Selectivity and effect of interference ions were successfully checked. The adsorbent is reusable for seven cycles and durable for six months. Freundlich isotherm is the best tted isotherm model and kinetic adsorption follows pseudo-rst order model. The Thermodynamic parameters clear that the adsorption is endothermic and spontaneous and it happens through physisorption. Seed oils were successfully applied as real samples.
Declarations Figure 1 FT-IR spectra of mGO and UreaLiCl-mGO