Modification of Magnetic Graphene Oxide by an Earth-Friendly Deep Eutectic Solvent to Preconcentrate Ultratrac Amounts of Pb(II) and Cd(II) in Legume Samples

Abstract

A novel magnetic solid-phase extraction adsorbent using deep eutectic solvent-coated magnetic graphene oxide (EgLiCl-mGO) was proposed for simultaneous preconcentration of Pb(II) and Cd(II). The nanocomposite was characterized by Fourier Transform Infrared Spectroscopy, X-ray diffractometry, and alternative gradient force magnetometer. Parameters that could affect the preconcentration recoveries of the target ions were investigated via the one-factor-at-a-time method. The optimum conditions are pH of 4 ± 0.5, EgLiCl-mGO amount of 1.0 × 10⁻² g, adsorption time of 5 min, eluent of HNO₃ (1 mL, 2 mol L⁻¹), and desorption time of one minute. The swelling property of the adsorbent versus pH was studied. The linearity of the dynamic range for Pb(II) (5.0 × 10⁻⁶–4.0 × 10⁻⁴ g L⁻¹) and Cd(II) (5.0 × 10⁻⁶–15 × 10⁻⁵ g L⁻¹) was recorded. The limits of detection were Pb(II): 1.2 × 10⁻⁶ g L⁻¹ and Cd(II): 47 × 10⁻⁸ g L⁻¹. The preconcentration factor of 50 was calculated for both ions and the relative standard deviations were 1.27% for Pb(II) and 0.94% for Cd(II). Reusability, effect of interference ions, selectivity, isotherm adsorption, kinetic adsorption, and thermodynamic adsorption were established. The adsorbent was successful at preconcentrating the ions in legumes.

1. Introduction

Deep eutectic solvent (DES) has received attention in the recent years. DES is an emerging class of safe materials related to ionic liquid (IL), which has abnormally deep melting point depression at the eutectic mixture of specific hydrogen bond donors (HBDs) and acceptors (HBAs). DESs are divided into five categories depending on the method of preparation [1,2]. Type I is composed of a quaternary ammonium salt and a metal chloride, Type II is a mixture of a quaternary ammonium salt and a metal chloride hydrate, Type III is a mixture of a quaternary ammonium salt and an HBD, Type IV consists of a metal chloride hydrate and an HBD, and Type V consists of only nonionic, molecular HBAs and HBDs. All types of DESs have high thermal stability, low volatility, low vapor pressure, and tunable polarity. Their preparation does not need any extra solvent so no purification step is needed. Therefore, they are applicable substitutions for volatile organic compounds (VOCs), used widely throughout research and industry with the aim of environmental remediation, especially magnetic solid-phase extraction (MSPE) [3,4,5].

MSPE is a developed generation of solid-phase extraction, which uses magnetic or magnetically modified adsorbents to extract analytes with high efficiency. This tactic overcomes some weak points in solid-phase extraction such as decreasing the time of preconcentration, sample loading, and filtration or centrifugation steps [6,7,8]. Magnetic graphene oxide is an example of an MSPE adsorbent. Graphene oxide (GO) is a fascinating new class of two-dimensional carbon nanostructures. It has high surface area, good chemical stability, and strong thermal stability. GO contains hydroxyl, epoxide, carboxyl, and carbonyl functional groups; additionally, its hydrophilic property increases its negative charge and dispersibility in aqueous solution to form a stable suspension. GO has strong interaction with magnetic nanoparticles; accordingly, mGO can be a good candidate to be applied MSPE as an adsorbent. GO can be functionalized with organic components via π–π interaction to control selective adsorption of analyte(s) [9,10,11,12]. DES is an example, which can be a modifier, disperser, and fictionalizer for mGO.

Many scientists have used DES to improve the extraction or preconcentration methods. For example, a DES of ethylene glycol and ammonium-based salt functionalized carbon nanotube to adsorb methyl orange from aqueous solution [13]. Additionally, a novel DES was applied as a modifier for graphene and graphene oxide to remove chlorophenols [14]. In additional to organic materials, DES can be a qualified candidate to adsorb cations such as heavy metals, because it has active negative sites and can grab cations via electrostatic force [15].

In this study, magnetic graphene oxide modified with DES (EgLiCl-mGO) was applied as an efficient nanoadsorbent for preconcentrating Pb(II) and Cd(II) simultaneously in legume samples. Pb(II) and Cd(II) are toxic elements and their accumulative characters cause serious problems for humans, plants, and animals [16]. According to specific properties of EgLiCl-mGO, it could be a highly applicable adsorbent. mGO facilities the preconcentration procedure because graphene oxide has a high surface area and magnetic properties enable easy and fast separation of EgLiCl-mGO [7,17,18,19,20]. The presence of DES provided a wide surface area with functional groups, which has a synergistic effect on adsorption efficiency. More importantly, EgLiCl is inexpensive, biodegradable, nontoxic, and easy to prepare. For the evaluation of the effective variables on the preconcentration efficiency, the one-at-a-time method was employed and pH, adsorption time, adsorbent amount, desorption time, and kind of eluent were optimized. Analytical figures of merit, interference effect of various ions, reusability, swelling behavior, adsorption isotherm, adsorption kinetics, and adsorption thermodynamics were reported. Finally, the target analytes were analyzed in the four different legumes.

2. Materials and Methods

2.1. Apparatus

Absorbance quantifications of Pb(II) and Cd(II) were conducted using a flame atomic absorption spectrometer (FAAS; Younglin Aas 8020 (http://youngincm.com, Gyeonggi, South Korea)) with a deuterium background correction system and an air–acetylene burner. All pH adjustments were conducted with a digital pH meter (Metrohm—827 (www.metrohm-ag.com, Herisau, Switzerland)). The instrument has a glass combination electrode. The X-ray powder diffraction pattern (XRD; Philips—PW1730 (www.panalytical.com, Eindhoven, The Netherland)) was obtained under the Cu-Kα radiation (1.2 kW). 2θ ranged from 10° to 80°, scan step and step time were 0.05° and 1 s, respectively. An alternative gradient force magnetometer (AGFM; Meghnatis Daghigh Kavir Company (https://nano.kashanu.ac.ir, Kashan, Iran)) measured magnetic properties in an applied magnetic field sweeping between ±10,000 Oe. Fourier transform infrared spectra (FT-IR; ABB Bomem MB100 (http://new.abb.com, Zürich, Switzerland)) were recorded over the range 400–4000 cm⁻¹.

2.2. Reagent and Solution

All the reagents were of analytical grade. The standards of Pb(II) and Cd(II) (1.0 × 10⁻³ g L⁻¹) were prepared using their nitrates salts. Flake graphite, P₂O₅, H₂SO₄, K₂S₂O₈, H₂SO₄, KMnO₄, NaNO₃, HCl, H₂O₂, FeCl₃·6H₂O, FeCl₂·4H₂O, NH₄OH, ethylene glycol (Eg), and LiCl were procured from Merck Company (www.merck.de, Darmstadt, Germany). Ultra-high purity water from a Milli-Q system was used to prepare sample solutions.

2.3. Synthesis

Graphene oxide (GO) was prepared using the modified Hummer’s method. A mixture of flake graphite (4.0 × 10⁻³ g), H₂SO₄ (12 mL), and P₂O₅ (8 × 10⁻³ g) was magnetically stirred for 6 h and then filtered. Afterwards, H₂SO₄ (12 mL) and K₂S₂O₈ (8.0 × 10⁻³ g) were added to the filtrate and was magnetically stirred for 6 h. The materials were cooled to room temperature, eluted by deionized water (300 mL), dried at room temperature, and then heated at 60 °C for 2 h. Pre-oxidized graphite powder (2.0 × 10⁻³ g), H₂SO₄ (92 mL), and KMnO₄ (12 × 10⁻³ g) were stirred together in an ice bath. After 15 min, NaNO₃ (2.0 × 10⁻³ g) was added to the materials and was stirred at room temperature for 2 h. Then deionized water (200 mL) was added and stirred for 15 min. H₂O₂ (30%, 10 mL) and distilled water (500 mL) were added. The materials were filtered and washed using HCl (10%) to reach the brown suspension, and then they were heated at 60 °C for 30 min. The cooled materials were sonicated in H₂O₂ (10%, 10 mL) for 5 min. The yellow-brown residual powder was washed with warm deionized water 3 times to remove the impurities. GO was dried at 60 °C [21,22].

Secondly, a magnetic carbonic nanocomposite was synthesized based on Massart’s method. Deionized water (100 mL), FeCl₃·6H₂O (2.5 g), FeCl₂·4H₂O (9.0 × 10⁻¹ g), and GO (1.0 × 10⁻¹ g) were sonicated for 15 min. Afterwards, under nitrogen atmosphere, NH₄OH (25%, 15 mL) was dripped to fix the pH~11. After stirring for 12 h, mGO was magnetically gathered, eluted using distilled water, and dried at 80 °C [23].

Finally, Eg (5.0 × 10⁻² g) and LiCl (5.0 × 10⁻² g) were mixed together at 60 °C. A colorless liquid shows a successful synthesis of DES. mGO (5.0 × 10⁻² g) was dispersed into the DES via one hour sonication. The final product (EgLiCl-mGO) was eluted using deionized water and then dried at 30 °C [24].

2.4. General Procedure

Pb(II) and Cd(II) (5.0 × 10⁻⁵ g L⁻¹) were dissolved in deionized water (50 mL). EgLiCl-mGO (1.0 × 10⁻² g) was added into the sample solution at the pH of 4 ± 0.5, followed by shaking for 5 min. EgLiCl-mGO was gathered magnetically. One-minute sonication in the presence of HNO₃ (1 mL, 2 mol L⁻¹) completed the process of preconcentration. Eluted ions were quantified by FAAS.

2.5. Sample Preparation

Four different legume samples including kidney bean, cowpea, pinto bean, and navy bean were procured from local stores in Tehran, Iran. The samples were cleaned with deionized water. They were dried to reach the stable weight, and then cooled to room temperature. The samples (1.0 × 10⁻¹ g) were transferred into a Teflon vessel containing HNO₃ (15 mL, 65%), and kept immersed for 48 h. The digestion heating steps are reported below: Firstly, the temperature was raised to 90 °C in 10 min and stayed for 5 min. Then, the temperature increased to 150 °C and remained for 10 min, followed by adding H₂O₂ (20 mL, 30%). Finally, the vessels were cooled down to room temperature. The solution in each vessel was transferred to a polyethylene volumetric flask and diluted to 100 mL. The pH of two separate 50 mL aliquots of digested samples was adjusted to 4 ± 0.5 using HNO₃ and NH₃ solutions. The general procedure was applied as mentioned in Section 2.4. Matrix spiking with standards of Pb(II) and Cd(II) (5.0 × 10⁻⁵) was employed to evaluate the effect of the matrix.

3. Results

3.1. Characterization

FT-IR spectra confirm that Eg-LiCl was loaded onto mGO completely (Figure 1). The FT-IR spectrum of mGO was drawn in the orange spectrum. The stretching O-H, carbonyl C=O, aromatic C=C, bending O-H, aromatic C-C, alkoxy C-O-C, stretching CH₂, and carbonyl C=O bonds are identified according to the peaks of 3380, 1730, 1630, 1360, 1220, 1050, 2880, and 1730 cm⁻¹, respectively. Peaks lower than 700 cm⁻¹ characterize bonds of Fe–O [25]. The blue spectrum is related to EgLiCl-mGO. The bands of 3430 and 1640 cm⁻¹ are associated O-H bonding. Alkane CH₂ stretching, CH₂ scissor, CH₂ wagging, and CH₂ twisting are characterized according to peaks around 2914, 1460, 1350, and 1258 cm⁻¹, respectively. The sorption peaks of Li-O and Fe-O were seen at the wavelengths of 1300 and 700 cm⁻¹, respectively [25]. Table S1 summarizes the data.

The X-ray diffraction (XRD) analysis of EgLiCl-mGO is patterned in Figure 2a, which has good agreement with JCPDS cards No. 01-075-0449, indicating magnetite Fe₃O₄ and JCPDS cards No. 01-079-1715 confirming the GO structure. Fe₃O₄ was characterized by peaks at 30.70, 35.97, 37.23, 43.74, 53. 94, 57.55, 63.32, and 67.14°. Additionally, peaks at 11.23, 21.64, 30.70, 35.97, 43.74, 53.94, 57.55, 63.32, and 75.08° are related to GO [26]. The process of loading EgLiCl onto mGO broadened the peaks of mGO. The crystal size of the total product was calculated using the Scherrer formula. It is about 6 nm.

The AGFM curve of EgLiCl-mGO is presented in Figure 2b. The amount of ±25 electromagnetic units (emu g⁻¹) is helpful for the rapid gathering of the adsorbent in the sample solution [27]. The remanent magnetization is almost zero, confirming the superparamagnetic property and the presence of Fe₃O₄ nanoparticles.

3.2. Optimization of the Method

3.2.1. Effect of pH

Alkaline or acidic condition of samples plays a crucial part for the preconcentration of heavy metals by solid adsorbents because it has a direct effect on Pb(II) and Cd(II) ion retention on the negative sites of EgLiCl-mGO. It may be possible for the cations to form coordination compounds with oxygen groups via electrostatic force. Recovery of the investigated ions on the EgLiCl-mGO was found Pb(II): 99.1 and Cd(II): 98.4 in pH of 4 ± 0.5. Quantitative recovery values were found in the pH range of 2–9. After pH~5, recovery values of the preconcentrated ions decreased because heavy metal ions were hydrolyzed with formation of metal hydroxides. Therefore, pH~4 was applied as the optimized pH value for further experiments. Additionally, in highly acidic pH, the negative active sites will be occupied by H⁺ sooner than the target cations.

3.2.2. Effect of Adsorption Time

A certain period of time must pass to reach a complete adsorption. Accordingly, a fixed amount of EgLiCl-mGO was added to the sample solutions of Pb(II) and Cd(II) (50 mL, 5.0 × 10⁻⁵ g L⁻¹) under the adsorption time, ranging from 1 to 30 min. The experiment results show that the performance of the method has increasing trend from 1 to 5 min, but after that there are slight differences in the recoveries. Therefore, the adsorption time of 5 min was selected for the next experiments.

3.2.3. Effect of Adsorbent Amount

An appropriate amount of adsorbent is an essential parameter in the MSPE method. On the one hand, this is effective for the usage amounts of materials; on the other hand, it provides sufficient active sites for adsorption of target analytes. In this context, the influence of EgLiCl-mGO amount on the simultaneous preconcentration efficiencies of Pb(II) and Cd(II) remained constant with the adsorbent amount ranging from 1.0 × 10⁻² to 5.0 × 10⁻² g. This indicates high affinity of EgLiCl-mGO toward these heavy metals. Therefore, 1.0 × 10⁻² g was selected for the next step of optimization.

3.2.4. Effect of Eluent Type

The influence of the eluent type on the simultaneous preconcentration efficiency of Pb(II) and Cd(II) by EgLiCl-mGO was compared in the presence of two acids (

Table 1. Results of different eluents (1 mL) on the preconcentration of Pb(II) and Cd(II) (n = 3).

Recovery (%)		Eluent
Cd(II)	Pb(II)
60.52 ± 0.03	41.23 ± 0.01	HNO3 (1 mol L−1)
98.14 ± 0.05	99.34 ± 0.07	HNO3 (2 mol L−1)
84.48 ± 0.11	76.15 ± 0.13	HNO3 (5 mol L−1)
14.96 ± 0.10	24.94 ± 0.14	HCl (1 mol L−1)
22.46 ± 0.15	11.16 ± 0.11	HCl (2 mol L−1)
5.63 ± 0.08	4.63 ± 0.11	HCl (5 mol L−1)

). The highest analytical yields for the target analyte were observed using HNO₃ (2 mol L⁻¹). Afterwards, the volume influence of HNO₃ on the extraction efficiency of the target analytes was studied in the range of 0.5–2 mL. The efficiency remained stable in all volumes. As one of the concerns in this article is using the minimum amounts of materials with the maximum performance, 1 mL was selected as the optimum volume of eluent.

3.2.5. Effect of Desorption Time

A completion of the desorption step depends on the desorption time. Insufficient desorption time does not let the eluent elute the adsorbed ions, resulting in poor analytical efficiency of the method. At the same time, it has a direct effect on the speed of preconcentration. Therefore, it is necessary to find the optimum desorption time. In the present study, the effect of desorption time on the method was studied in the range of 1–5 min and 1 min of shaking is sufficient for desorption of Pb(II) and Cd(II).

3.3. Swelling Behavior of EgLiCl-mGO

The swelling behavior of the adsorbent in different pH was investigated through the recording of the swelling ratio of EgLiCl-mGO in aqueous solution (50 mL) of Pb(II) (5.0 × 10⁻⁵ g L⁻¹) and Cd(II) (5.0 × 10⁻⁵ g L⁻¹). A certain mass of the adsorbent (1.0 × 10⁻² g) was shaken into sample solution at different pH (ranging from 2 to 9) at 25 °C for 5 min. Then, the weights of the swelled adsorbent were recorded after decanting the water. The swelling ratio (SR) of the fibers was determined using Equation (1):

$$SR=\frac{W_s-W_d}{W_d}$$

(1)

where W_s and W_d represent the weight of the wet adsorbent in water and that of dry adsorbent. Additionally, the pH effect of the sample solution on the swelling ratio was explored by recording SR in different pH ranging from 2 to 9 (

Table 2. Swelling behavior of EgLiCl-mGO to adsorb Pb(II) and Cd(II) at different pH.

pH	Wd (×10−2 g)	Ws (×10−2 g)	SD
2	1.0	1.7	0.7
3	1.0	6.3	5.3
4	1.0	9.3	8.3
5	1.0	8.2	7.2
6	1.0	7.9	6.9
7	1.0	6.0	5.0
8	1.0	3.1	2.1
9	1.0	2.1	1.1

). By increasing pH from 2 to 4, the swelling has an increasing trend but after that decreased significantly [28,29,30]. In all pH, EgLiCl-mGO adsorbed molecules of water via electrostatic interaction but in pH~4 the adsorption of Pb(II) and Cd(II) reached their maximum. The ion adsorption causes swelling of the adsorbent, illustrating the ability of EgLiCl-mGO to adsorb heavy metals.

3.4. Reusability of EgLiCl-mGO

First, EgLiCl-mGO (1.0 × 10⁻² g) was shaken in an aqueous solution (50 mL) of Pb(II) (5.0 × 10⁻⁵ g L⁻¹) and Cd(II) (5.0 × 10⁻⁵ g L⁻¹) and the amounts of heavy metals ions in the eluent were determined using FAAS. Five consecutive cycles were completed according to the procedure in Section 2.4. The metal ion adsorption efficiencies decreased after the second cycle. The method is reusable for two times. This decrease is related to negative effect of eluent onto EgLiCl-mGO. Eluent impacts on EgLiCl and affect the interaction among the components.

3.5. Interference Effect of Various Ions and Selectivity of EgLiCl-mGO

The study of interreference effects on the suggested method is important, because the adsorption of Pb(II) and Cd(II) ions may be influenced by other cations and anions; consequently, the performance of the method decreases [31]. Adsorption of the target ions by EgLiCl-mGO was conducted in the presence of higher amounts of other ions with respect to their applications in real samples. The ±5% was considered as the maximum tolerance limit and values smaller than this amount were acceptable.

Table 3. The interfering effect of various ions on preconcentrating Pb(II) and Cd(II) (n = 3).

Ions	Ratio of Coexisting Ions	Recovery (%)
		Pb(II)	Cd(II)
Na(I)	10,000	96.03 ± 0.12	98.14 ± 0.09
K(I)	10,000	98.12 ± 0.27	97.43 ± 0.11
Li(I)	800	99.17 ± 0.14	99.14 ± 0.08
Ca(II)	500	100.56 ± 0.21	99.17 ± 0.05
Mg(II)	300	98.08 ± 0.18	96.13 ± 0.04
Cr(II)	200	97.03 ± 0.24	99.94 ± 0.18
Pd(II)	800	97.26 ± 0.19	98.32 ± 0.11
Zn(II)	250	102.45 ± 0.11	100.32 ± 0.10
Al(III)	700	99.69 ± 0.27	98.71 ± 0.07
Mn(II)	500	97.34 ± 0.18	96.52 ± 0.04
Co(II)	100	98.98 ± 0.35	101.43 ± 0.01
Cu(II)	100	96.15 ± 0.28	96.83 ± 0.04
Ni(II)	100	95.28 ± 0.17	99.54 ± 0.9
SO42−	1000	95.35 ± 0.31	98.18 ± 0.13
Cl−	1000	97.53 ± 0.24	100.32 ± 0.15
NO3−	1000	99.18 ± 0.19	99.54 ± 0.09
CO32−	1000	101.37 ± 0.20	100.91 ± 0.11

summarizes the data.

Under the optimum conditions, the selectivity of the method was studied. According to the data, the preconcentration efficiencies of Pb(II) and Cd(II) were 99% and 98%, respectively; however, the preconcentrating efficiencies of Cr(II), Cu(II), Mn(II), and Pd(II) were 32%, 85%, 17%, and 76%, respectively. This strength is related to selective interaction between target ions and specific active sites on the surface of EgLiCl-mGO. Pb(II) and Cd(II) have a smaller ionic radius so they have a greater charge density; consequently, they occupy the active sites of EgLiCl-mGO sooner and more strongly.

3.6. Sample Analysis

The accuracy of EgLiCl-mGO for simultaneous preconcentration of Pb(II) and Cd(II) in real samples was assessed through the spike method (5.0 × 10⁻⁵ g L⁻¹). Four different legumes including kidney bean, cowpea, pinto bean, and navy bean were purchased from local stores in Tehran, Iran. The sample preparation is explained in the section of sample preparation in detail, followed by applying the optimum procedure regarding Section 2.4. The results are presented in

Table 4. Analytical results of Pb(II) and Cd(II) quantified by EgLiCl-mGO (n = 3).

Sample	Spiked (g L−1)	Found	Recovery (%)	Found	Recovery (%)
		Cd(II)		Pb(II)
Kidney bean	0	1.84	-	2.46	-
	5.0 × 10−5	52.91	102.14	54.26	103.43
Cowpea	0	1.34	-	1.76	-
	5.0 × 10−5	49.65	96.97	49.64	97.79
Pinto bean	0	3.11	-	2.95	-
	5.0 × 10−5	51.98	97.87	50.75	95.84
Navy bean	0	0.95	-	4.76	-
	5.0 × 10−5	52.04	102.15	54.38	99.30

. The recovery values for the analyte ions were satisfactorily reasonable in the range 95–101%. The amounts of Pb(II) and Cd(II) were quantified in both the spiked and unspiked samples regarding Equation (2). The determination of ions was conducted using a FAAS.

$$R\%=\frac{C_1-C_2}{C_3}\times 100$$

(2)

where C₁, C₂, C₃, and R% are spiked portion, unspiked portion, the amounts of the ions, and relative recovery, respectively.

3.7. Analytical Figures of Merit

Analytical figures of merit for the suggested procedure under the optimum conditions (Section 2.4) were determined from results of the analyses. The linear calibration equations for Pb(II) and Cd(II) were A = 0.0011C + 0.0047 (R² of 0.988) and A = 0.0112C − 0.0172 (R² of 0.998), respectively. Linearity of dynamic range (LDR) for Pb(II) (5.0 × 10⁻⁶–4.0 × 10⁻⁴ g L⁻¹) and Cd(II) (5.0 × 10⁻⁶–15 × 10⁻⁵ g L⁻¹) were recorded. In these equations, A is the absorbance of ions and C expresses as their concentrations in initial sample solutions. The limits of detection (LOD), expressed as 3S_b/m (S_b is blank standard deviation and m is slope of the calibration plot), were found as Pb(II) 1.2 × 10⁻⁶ g L⁻¹ and Cd(II) 47 × 10⁻⁸ g L⁻¹ for five measurements of the blank. The limits of quantification (LOQ) were Pb(II) 3.9 × 10⁻⁶ g L⁻¹ and Cd(II) 1.5 × 10⁻⁶ g L⁻¹. LOQ was defined as 3.33 LOD. The preconcentration factor (PF) of 50 was calculated as the concentrations of ions before and after preconcentration. The reproducibility of the preconcentration method (relative standard deviation (RSD)) was determined by performing five experiments from solutions containing Pb(II) and Cd(II) (50 mL, 5.0 × 10⁻⁵ g L⁻¹). The results showed 1.27% for Pb(II) and 0.94% for Cd(II).

Table 5. Comparison of the proposed technique with several recent methods for simultaneous determination of heavy metals (FAAS was used for all detections).

Method	Amount (g)	LOD (g L−1)	RSD (%)	PF	LDR (g L−1)	Reference
SPE	-	6 × 10−6–12 × 10−6	1.0–2.2	100	0.5 × 10−6–100 × 10−6	[32]
MSPE	5.0 × 10−2	0.06 × 10−6–0.76 × 10−6	0.18–3.10	150	0.1 × 10−6–200 × 10−6	[33]
LLE	-	0.38 × 10−6–0.42 × 10−6	3.6–5.2	-	1.0 × 10−6–40 × 10−6	[34]
SPE	4.0 × 10−3	2.74 × 10−6–3.10 × 10−6	3.2–3.5	50	-	[35]
SPE	-	0.031 × 10−6–0.043 × 10−6	2.44–4.68		0.25 × 10−6–25 × 10−6	[36]
MSPE	1.0 × 10−2	47 × 10−8–1.2 × 10−6	0.94–1.27	50	5.0 × 10−6–4.0 × 10−4 g L−1	This work

shows a comparison among EgLiCl-mGO with some new adsorbents. EgLiCl-mGO has an important development in MSPE including decreasing adsorbent amount and RSD as well as widening LDR. Additionally, EgLiCl-mGO was performed in semi-neutral pH, reducing the usage of materials to pH adjustment. More importantly, EgLiCl-mGO is a green, safe, and earth-friendly adsorbent and its preparation is fast without consumption of any dangerous material. The preconcentration time is about six minutes, which is another advantage of this method.

3.8. Adsorption Isotherm

The isotherm study shows the adsorption behavior of a system such as homogenous or heterogenous distribution of the analyte onto the surface of the adsorbent. In this article, five isotherm models were applied to investigate the adsorption of Pb(II) and Cd(II) on the surface of EgLiCl-mGO. This study was conducted in the presence of EgLiCl-mGO (1.0 × 10⁻² g) in the sample solution containing the analytes (50 mL) during 5 min shaking at room temperature. According to Figure S1, increasing the initial concentration of ions has a direct effect on adsorption capacity (q_e) of EgLiCl-mGO, because by increasing C_e, q_e increases. q_e is calculated using the following equation.

$$q_e=\frac{(C_0-C_e)}{W}\times V$$

(3)

C₀ is the initial concentration of the analyte, C_e is the equilibrium concentration of the analyte, V is the volume of the sample solution, and W is the amount of the adsorbent [37,38].

Langmuir is a monolayer adsorption model. It shows that the adsorbate molecules are adsorbed onto the homogeneous surface of the adsorbent. After the first layer of the analyte, no more analyte will be adsorbed onto the surface of the adsorbent. Equation (4) shows the linear form of this model.

$$\frac{1}{q_e}=\frac{1}{q_0{K}_L{C}_e}+\frac{1}{q_0}$$

(4)

C_e is the equilibrium concentration of heavy metals adsorbed by EgLiCl-mGO, q_e is the retention capacity of analytes, K_L is the Langmuir constant, showing the binding energy between the adsorbate and adsorbent; q_max is the maximum adsorption capacity of the adsorbent. The straight line was obtained when 1/q_e was plotted against 1/C_e as shown in Figure S2. The slope and intercepts show the q_max and K_L, respectively [38,39,40].

The Freundlich model is a multilayer adsorption model. The linear form of the Freundlich model is written in Equation (5).

$$\ln q_e=\ln k_f+\frac{1}{n}\ln C_e$$

(5)

where K_f is the Freundlich constant, expressing the capacity, C_e defines the adsorbate concentrations of analytes, q_e is the quantity of ions at equilibrium, and n is the Freundlich exponent. K_f and n were determined from plot lnq_e vs. lnC_e (Figure S3). The n value shows the nature of adsorption (1/n < 1: normal process), (1/n > 1: cooperative process) [38,39,40].

The Temkin isotherm depicts the interaction between the adsorbate and the adsorbent and causes a linear decrease in the adsorption energy with surface coverage of the adsorbent. Equation (6) shows the linear form of this model.

$$q_e=\frac{RT}{B}\ln A_T+\frac{RT}{B}\ln C_e$$

(6)

where A_T is the Temkin isotherm equilibrium binding constant, R of 8.314 × 10³ J mol⁻¹ K⁻¹ is the universal gas constant, T is the absolute temperature, and B is the model constant and shows the heat of absorption [40,41]. Both B and A_T can be extracted from the plot q_e vs. lnC_e (Figure S4).

The Halsey isotherm model expresses the multilayer adsorption of the analyte over the heterosporous adsorbent. Equation (7) shows the linear form for the model.

$$\ln q_e=(\frac{1}{n_H})\ln K_H-(\frac{1}{n_H})\ln C_e$$

(7)

The Halsey isotherm model constant and Halsey isotherm model exponent are represented as K_H and n_H, respectively [38,40,41]. This model is plotted in Figure S5.

The Elovich model is based on multilayer adsorption, assuming exponential increasing trend in the adsorption sites as adsorption takes place. It is given by the equation below:

$$\ln \left(\frac{q_e}{C_e}\right)=\ln (K_e{q}_m)-\frac{q_e}{q_m}$$

(8)

K_e and q_m are the Elovich constants, representing the initial adsorption rate and maximum adsorption capacity, respectively [38,40,41]. Figure S6 shows the model.

Among the five mentioned models, Pb(II) follows the Freundlich model to be adsorbed onto EgLiCl-mGO, but Langmuir is fitted with the adsorption of Cd(II) onto EgLiCl-mGO. Therefore, the adsorption of Pb(II) onto the surface of EgLiCl-mGO is multilayered and a different layer of Pb(II) can be adsorbed onto each other but the adsorption of Cd(II) onto the surface of EgLiCl-mGO is the monolayer and after the first layer of Cd(II) no more Cd(II) is adsorbed. Isotherm parameters of each model are summarized in

Table 6. Adsorption isotherms for Pb(II) and Cd(II) onto EgLiCl-mGO.

Model	Pb(II)		Cd(II)
	Parameters	R2	Parameters	R2
Langmuir	q0 = 75.18 KL = 2.83	0.99	q0 = 50 KL = 1.45	0.90
Freundlich	n = 4.52 Kf = 45.36	0.89	n = 2.48 Kf = 24.77	0.99
Temkin	B = 233.31 AT = 100.48	0.92	B = 178.38 AT = 7.02	0.93
Elovich	qm = 1.51 KE = 727.18	0.86	qm = 25.1 KE = 2.76	0.90
Halsey	n = −4.52 KH = 31.6 × 106	0.86	n = −2.4 KH = 2.9 × 103	0.97

. R² shows the best model for each of analyte. All the figures (Figures S1–S6) are plotted in the Electronic Supplementary Materials (ESMs).

3.9. Adsorption Kinetics

The rate and the characteristics of the adsorption of target analytes onto EgLiCl-mGO were investigated via pseudo-first-order and pseudo-second-order models. The samples of heavy metals (50 mL, 20 mg L⁻¹) were prepared and EgLiCl-mGO (1.0 × 10⁻² g L⁻¹) was added. The adsorption of ions was followed up at different time points, ranging from 1 to 60 min at room temperature.

The pseudo-first-order kinetic modeling is provided in the following equation:

$$\ln (q_e-q_t)=\ln q_{\max }-k_{1}t$$

(9)

q_t (mg g⁻¹) and k₁ (min⁻¹) are the amounts of adsorbed ions onto EgLiCl-mGO at several times and the pseudo-first-order kinetic, respectively [42,43,44,45]. Figure S7 shows the plot.

The pseudo-second-order kinetic is plotted according to Equation (10).

$$\frac{t}{q_t}=\frac{1}{k_2\times q{}_{cal}{}^2}+\frac{1}{q_{cal}}\times t$$

(10)

where q_t (mg g⁻¹) and k₂ (g mg⁻¹ min⁻¹) are the amount of adsorbed analyte onto the surface of solid adsorbent at various times and the rate constant of the pseudo-second-order kinetic, respectively [42,43]. Figure S8 shows the plot.

Figures S7 and S8 are presented in the Electronic Supplementary Materials (ESMs). The R² of the pseudo-first-order model was in the scope of 0.99 for both ions so the adsorption rate-controlling step was performed by occupying empty adsorption sites of EgLiCl-mGO by Pb(II) and Cd(II) (

Table 7. Comparison of kinetic models for Pb(II) and Cd(II) adsorption onto EgLiCl-mGO.

Analyte	Model	Parameters	Values	R2
Pb(II)	Pseudo first order	k1 (min−1) qe (mg g−1)	0.03 95.40	0.99
	Pseudo second order	k2 (g mg−1 min−1) qe (mg g−1)	4.04 × 10−4 111.11	0.95
Cd(II)	Pseudo first order	k1 (min−1) qe (mg g−1)	0.023 67.46	0.99
	Pseudo second order	k2 (g mg−1 min−1) qe (mg g−1)	1.7 × 10−3 88.49	0.98

) [44,45,46].

3.10. Adsorption Thermodynamics

The adsorption process of heavy metals onto solid adsorbents mainly takes places via electro-static forces so it is exothermic (ΔH < 0) or endothermic (ΔH > 0) in nature. The value of ΔH is the main parameter used to distinguish chemisorption from physisorption [47]. K_d is the equilibrium constant, showing free Gibbs energy (ΔG); consequently, the enthalpy (ΔH) and the entropy ΔS of adsorption could be calculated using the following equations:

$$\Delta G=-RT\ln k_d$$

(11)

$$\Delta G=\Delta H-T\Delta S$$

(12)

where T is the absolute temperature and R is the universal gas constant (8.314 J mol⁻¹ K⁻¹). The plots of ln(K_d) vs. 1/T for Pb(II) and Cd(II) are shown, respectively, in Figure S9a,b in the Electronic Supplementary Materials (ESMs). The negative values of ∆G indicate the spontaneous nature of Pb(II) and Cd(II) adsorption onto EgLiCl-mGO. The positive values of ∆H demonstrate the endothermic character of the adsorption process, and ΔH more than 40 kJ mol⁻¹ shows chemisorption and ΔH lower than 40 is related to chemisorption. The positive values of ∆S indicate the increase in disorder at the solid–liquid interface as well as more agitation in the sample solution due to releasing water molecules that grab the ions [48,49,50]. The obtained results are shown in

Table 8. Thermodynamic parameters for the Pb(II) and Cd(II) adsorption onto EgLiCl-mGO.

T	Pb(II)			Cd(II)
	Ce (mg L−1)	qe (mg g−1)	ΔG (J mol−1)	Ce (mg L−1)	qe (mg g−1)	ΔG (J mol−1)
273	28.65	6.75	3581	25.43	22.85	265
283	26.43	17.85	972	23.45	32.75	−827
298	22.91	35.45	−1081	20.85	45.75	−1946
318	14.65	76.75	−4103	17.76	61.2	−3065
343	7.13	114.35	−6875	12.21	88.95	−4920

.

4. Discussion

In the present study, a novel nanocomposite (EgLiCl-mGO) was successfully synthesized and applied as an efficient adsorbent in the MSPE procedure of Pb(II) and Cd(II) ions. The nanocomposite combining DES had the advantage of superparamagnetism, low environmental pollution, good water dispersibility, rapid extraction, and reproducibility; in the meantime, it showed excellent selectivity for Pb(II) and Cd(II). Moreover, its structural and magnetic properties were confirmed by FT-IR, VSM, and XRD. For the MSPE procedure, the affecting parameters on adsorption and desorption steps were optimized via the one-at-a-time method. LOD, LOQ, PF, EF, RSD, selectivity, effect of interference ions, selectivity, swelling behavior, isotherm adsorption, kinetic adsorption, and thermodynamic adsorption were investigated in detail. The technique was applied successfully to preconcentrate Pb(II) and Cd(II) simultaneously in four legume samples.

5. Conclusions

This article confirmed that EgLiCl-mGO is a qualified adsorbent to simultaneously preconcentrate ultratrace amounts of Pb(II) and Cd(II) in legumes. In additional to acceptable analytical results, this adsorbent is easy to prepare without consumption of dangerous and harmful reagents. The EgLiCl-mGO is an energy efficient and environmentally friendly adsorbent.