Nickel separation from human blood samples based on amine and amide functionalized magnetic graphene oxide nano structure by dispersive sonication micro solid phase extraction

6 Anal. Method Environ. Chem. J. 3 (1) (2020) 5-16 Toxic effects of oral exposure to nickel usually involve the kidneys with some evidence from animal studies showing a possible developmental/ reproductive toxicity effect (ATSDR). Normal range for Ni in healthy peoples is 0.2 μgL-1 in serum and less than 3.0 μgL-1 in human urine. A national health and nutritional examination survey (NHNES) of hair found mean nickel levels of 0.39 mg L-1, with 10% of the population having levels less than1.5 mg L-1 [4,5]. The vary techniques was used for measurement of nickel (Ni2+) in human bodies such as serum and blood samples. Due to previous studies, the flame atomic absorption spectrometry (F-AAS) and electrothermal atomic absorption spectrometry (ET-AAS) were reported for analysis of heavy metals such nickel (Ni) which was suitable for the determination in biological matrixes [6-9] . Occasionally, the atom trapping flame atomic absorption spectrometry (AT-FAAS) and fluorescence spectrometry (XRF) also was used for heavy metal determaintion such as nickel in biological samples [10, 11]. Also, different methods such as, polarography [12], inductively coupled plasma-optical emission spectrometry (ICP OES, inductively coupled plasma-mass spectrometry (ICP-MS) [13, 14], and spectrophotometry [15] were used for nickel analysis in human samples [16]. But as difficulty matrixes in human biological samples such as blood or serum, the sample preparation is required to separation nickel ions from liquid phases. The different sample preparation technology exist for nickel extraction from blood samples such as liquid– liquid extraction (LLE) [17], solid-phase extraction (SPE) and functionalized magnetic-SPE [18], dispersive solvent by liquid–liquid microextraction method (DLLME) [19], phase extraction based on cloud point (CPE) [20], dispersive solid phase microextraction (D-SPME) [21], ultrasoundassisted solid phase extraction (USA-SPE) [22]. Recently, the dispersive sorbent in the liquid phase was presented as micro SPE (D-μ-SPE) for separation/determination of nickel in water and biological samples [23]. Some advantages such as high extraction efficiency, simple usability, fast and low time caused to select as a favorite technique for metal extraction. The sorbent characterizations are a main factor which was effected on heavy metal extraction by the D-μ-SPE procedure. The selective of favorite nanosturactures improved recovery. The different nanosorbent such as, carbon nanotubes (CNTs), graphene / graphene oxide sheets (NG, NGO) and silica (MSN) were used for extraction of Ni in blood samples [24-26]. Between them, the NGO was reported as efficient sorbents for metal extraction due to their surface properties. In this study, a novel sorbent based on Fe3O4supported amine/amide-functionalized graphene oxide (Fe3O4@A/A-GO) was used for separation/ speciation Ni(II) from human blood samples by dispersive sonication micro solid phase extraction (DS-μ-SPE) at pH=8.0. The method was developed in blood samples by ET-AAS. 2. Experimental 2.1. Instrumental The graphite furnace atomic absorption spectrophotometer (GF-AAS, 932 GBC, Aus) was used for nickel determination in blood samples with the Avanta software. The linear range of 3.0-150 μg L-1 (peak Area, Abs=1.91) was selected for Ni in optimized light. The current and wavelength of HCL lamp was adjusted 3.0 mA and 228.8 nm, respectively. The auto-sampler of spectrophotometer (Pal 3000) was used for micro injection (1μL) of sample volumes to furnace tube after adjusted injector. The pH of the sample was digitally calculated by Metrohm pH meter (Swiss). Fourier transform infrared (FT‐IR) spectra were recorded from KBr pellets using a Perkin Elmer Spectrum 65 FT‐IR spectrophotometer. Powder X-ray diffraction (XRD) was conducted on a Panalytical X’Pert PRO X-ray diffractometer. Scanning electron microscopy (SEM) images were obtained using a Tescan Mira-3 Field Emission Gun Scanning Electron Microscope (SEM). Magnetic 7 Nickel separation in blood by Fe3O4@A/A-GO Anne Trégouët, et al property of the catalyst sample was measured with a vibrating sample magnetometer (VSM) model LBKFB, Meghnatis Daghigh Kavir Co, Iran, at room temperature 2.2. Reagents Chemicals including natural flake graphite (325 mesh, 99.95%), were purchased from Merck chemical company and used as received. The standard stock solutions (1000 mg L-1) of Ni (II), acetone, acetate, HNO3, NaOH, HCl and other reagents were purchased from Merck (Darmstadt, Germany). Ultra-pure deionized water (DW) purchased from Milli-Q plus water purification system (USA). The standard and experimental solutions of Ni2+ (0.1, 0.2.0.5, 1.0, 3, 5.9 and 7.0 μg L-1) were prepared daily by appropriate dilution of the stock solutions with DW. The pH of samples were used with appropriate buffer solutions including sodium acetate for pH 3.5–5.6, sodium phosphate for pH of 5.8–8.0, and ammonium chloride for pH 8–10. All the laboratory glassware and plastic tubes were cleaned by 5% (v/v) HNO3 for at least 12 h and then washed many times with DW and dried in oven prior to use. All reagents for synthesis of Fe3O4@A/A-GO prepared by RIPI Company. 2.3. Synthesis of graphene oxide (GO) Graphite black powder was oxidized to GO following the modified Hummers method in the several steps [27-29]. Fore pre oxidation of graphite powder, 250 mL of H2SO4 was added to 5 g of graphite powder and the resulting mixture was stirred for 24 h. After 24 h, 30 g of KMnO4 was added to the mixture stirring for 72 h at 50 °C. Next, a solution of 45 mL of H2O2 (30%) in 400 mL of deionized water was added to the mixture after which the brown color of the mixture turned into bright yellow. The GO dark solution was centrifuged and washed with deionized water and 10% HCl solution, and dried at 65 °C. 2.4. Preparation of nanomagnetic Fe3O4supported Amine/Amide-functionalized graphene oxide (Fe3O4@A/A-GO) 2.4.1. Acyl-chlorination of graphene oxide In the first step, thionyl chloride 60 mL was added GO (0.3 g) and stirred in at 70°C for 24 h. After acyl-chlorination reaction, the Acyl-chlorination of GO was washed with THF four times and the precipitated dried at 65°C [30]. 2.4.1 Functionalization of graphene oxide with Ammonia A total of 1 g of Ammonia was added to Acylchlorinated GO (0.3 g). Then, the mixture was refluxed for 72 h at 120°C under argon condition. Finally, the mixture reaction washed with DI water. A dark powder of Amine/Amide-functionalized graphene oxide was obtained [31]. 2.4.2 Nano magnetization of Amine/Amide-funcf tionalized graphene oxide The Fe3O4-supported Amine/Amide-functionalized graphene oxide (Fe3O4@A/A-GO) nanoparticles were synthesized by co-precipitation of FeCl2·4H2O and FeCl3·6H2O, in the presence of Amine/Amide-grafted graphene oxide. First, a solution of FeCl2·4H2O and FeCl3·6H2O was prepared with a molar ratio of 2:1. The weight ratio of GO to FeCl3 in the nano composite was 1 per 20 (mGO: mFeCl3 = 1:20). Afterward, 10 mg of Amine/Amide-grafted graphene oxide in 15 mL of DI water was ultrasonicated for 20 min. 12.5 mL solution of FeCl2·4H2O (125 mg) and FeCl3·6H2O (200 mg) in deionized water (10 mL) was added to the reaction mixture at room temperature. In order to raise pH value to 11, an aqua solution of 30% ammonia was added in the reaction mixture at 70 °C. Then, the reaction mixture was allowed to cool to room temperature and Fe3O4@A/A-GO washed five times with DI H2O dried at 70 °C [32]. 2.5 Analytical Procedure In proposed procedure, 10 mL of blood and Nickel separation in blood by Fe3O4@A/A-GO Anne Trégouët, et al 8 Anal. Method Environ. Chem. J. 3 (1) (2020) 5-16 standard solutions was prepared by buffer solution. The whole blood samples diluted with DW (1:1) before procedure. By DS-μ-SPE procedure, the pH of the standard solution containing 0.2 -7.0 μg L−1 of nickel was adjusted up to 8 and then 0.01 g of Fe3O4@A/A-GO as adsorbent was added to samples. The standard / blood samples was shaked for 5.0 min at room temperature and Ni ions physically adsorbed on surface of Fe3O4@A/AGO and chemically extracted by amine and amide covalence bonding at pH=8. Then, the Fe3O4@A/ A-GO separated from liquid phase by magnetic accessory. Finally, the Ni ions back-extraction with nitric acid (0.2 M) from Fe3O4@A/A-GO and after dilution up to 0.5 mL with DW was determined by ET-AAS. In optimized conditions, the recovery of physically and chemically adsorption by Fe3O4@A/ A-GO was almost obtained 25.3% and 97.8%, respectively. The extraction efficiency of proposed method based on Fe3O4@A/A-GO was calculated by equation 1. The CA is the stock concentrations of nickel and CB is the remain concentration of Ni(II) after procedure (n=10, Eq. 1). Recovery of extraction=(CA-CB)/CA×100 (Eq.1) 3. Results and Discussion 3.1. Characterization of the nano Fe3O4@A/A-GO The Fe3O4-supported Amine/Amide-functionalized graphene oxide nanomagnetic (Fe3O4@A/A-GO) was synthesized according to the synthetic route shown in Figure 1. Firstly, graphite was oxidized to GO by the modified Hummers method [27-29]. Afterward, the GO was amination and amidation with Ammonia by according to our previously reported procedure [30, 31]. Finally, the resulted Amine/Amide-functionalized graphene oxide (Fe3O4@A/A-GO) was nano-magnetized by co‐ precipitation of ferrous (Fe2+) and ferric (Fe3+) ions in the presence of A/A-GO to afford the target adsorbent Fe3O4@A/A-GO [32]. Figure 2 shows the FT-IR spectra of GO and Fe3O4@A/A-GO. The broad peak in the range between 2600-3500 cm-1 in the IR spectra of these compounds is related to the (O-H stretching) Fig 1. Synthetic route of Fe3O4@A/A-GO Scheme 2. Synthetic route of Fe3O4@A/A-GO 2.5. Analytical Procedure In proposed procedure, 10 mL of blood and standard solutions was prepared by buffer solution. The whole blood samples diluted with DW (1:1) before procedure. By DS-μ-SPE procedure, the pH of the standard solution containing 0.2 -7.0 μg L of nickel was adjusted up to 8 and then 0.01 g of Fe3O4@A/A-GO as adsorbent was added to samples. The standard / blood samples was shacked for 5.0 min at room temperature and Ni ions physically adsorbed on surface of Fe3O4@A/A-GO and chemically extracted by amine and amide covalence bonding at pH=8. Then, the Fe3O4@A/A-GO separated from liquid phase by magnetic accessory. Finally, The Ni ions back-extraction with nitric acid (0.2 M) from Fe3O4@A/A-GO and after dilution up to 0.5 mL with DW was determined by ET-AAS. In optimized conditions, the recovery of physically and chemically adsorption by Fe3O4@A/A-GO was almost obtained 25.3% and 97.8%, respectively. The extraction efficiency of proposed method based on Fe3O4@A/A-GO was calculated by equation 1. The CA is the stock concentrations of nickel and CB is the remain concentration of Ni(II) after procedure (n=10, Eq. 1). Recovery of extraction = (CA-CB)/CA×100 (Eq.1) 3. Results and Discussion 3.1. Characterization of the nano Fe3O4@A/A-GO The Fe3O4-supported Amine/Amide-functionalized graphene oxide nanomagnetic (Fe3O4@A/A-GO) was synthesized according to the synthetic route shown in Scheme 2. Firstly, graphite was oxidized to GO by the modified Hummers method [27-29]. Afterward, the GO was amination and amidation with Ammonia 9 Nickel separation in blood by Fe3O4@A/A-GO Anne Trégouët, et al vibration of carboxylic and enolic functionalities [35]. The peaks at 3420, 1719, 1621, and 1060 cm-1 shown in the spectra of GO and Fe3O4@A/AGO are ascribed to the (C-O stretching), (C=C stretching), (C=O stretching), and (O-H stretching) respectively [33, 34]. Also, The absorption bands at 3360, 3181, 1630, and 1234 cm-1 shown in the spectra of GO and amination GO are ascribed to the stretching bands ν(O-H), ν(N-H), ν(C=O), and ν(C-N) respectively. In the spectrum of Fe3O4@A/ A-GO, the peaks observed at around 628 and 583 cm-1 are related to the Fe–O stretching vibration [37-39]. These results prove that the successful Amination and amidation of GO and synthesis of Fe3O4@A/A-GO. The XRD patterns of GO, and Fe3O4@A/A-GO are demonstrated in Figure 3 (a,b). GO has the two main peak at 2θ = 11.5°, 42.58o are related to the diffraction planes of (002) and (100) respectively, which can be observed in the XRD patterns of both GO and Fe3O4@A/A-GO [40-42]. As shown in Figure 3 (b), the peaks at 2θ = 24° and 42.58o It is evident from the XRD pattern of Fe3O4@A/AGO that was proved the presence of GO [42]. The main peaks at 2θ° = 30.12, 35.45, 43.07, 56.97, 62.47 in XRD pattern of Fe3O4@A/A-GO are in good accordance with the standard XRD data of magnetite Fe3O4. The main diffraction peaks at by according to our previously reported procedure [30, 31]. Finally, the resulted Amine/Amidefunctionalized graphene oxide (Fe3O4@A/A-GO) was nano-magnetized by co‐precipitation of ferrous (Fe) and ferric (Fe) ions in the presence of A/A-GO to afford the target adsorbent Fe3O4@A/A-GO [32]. Fig. 2 shows the FT-IR spectra of GO and Fe3O4@A/A-GO. The broad peak in the range between 26003500 cm in the IR spectra of these compounds is related to the (O-H stretching) vibration of carboxylic and enolic functionalities [35]. The peaks at 3420, 1719, 1621, and 1060 cm shown in the spectra of GO and Fe3O4@A/A-GO are ascribed to the (C-O stretching), (C=C stretching), (C=O stretching), and (O-H stretching) respectively [33, 34]. Also, The absorption bands at 3360, 3181, 1630, and 1234 cm-1 shown in the spectra of GO and amination GO are ascribed to the stretching bands ν(O-H), ν(N-H), ν(C=O), and ν(C-N) respectively. In the spectrum of Fe3O4@A/A-GO, the peaks observed at around 628 and 583 cm are related to the Fe–O stretching vibration [37-39]. These results prove that the successful Amination and amidation of GO and synthesis of Fe3O4@A/A-GO. Fig. 2. FT-IR spectra of (a) GO, and (b) Fe3O4@A/A-GO The XRD patterns of GO, and Fe3O4@A/A-GO are demonstrated in Fig. 4. GO has the two main peak at 2θ = 11.5°, 42.58 are related to the diffraction planes of (002) and (100) respectively, which can be observed in the XRD patterns of both GO and Fe3O4@A/A-GO [40-42]. As hown in Fig. 4 (b), the peaks at 2θ = 24° and 42.58 It is evident from the XRD pattern of Fe3O4@A/A-GO that was proved the presence of GO [42]. The main peaks at 2θ° = 30.12, 35.45, 43.07, 56.97, 62.47 in XRD pattern of Fe3O4@A/A-GO are in good accordance with the standard XRD data of magnetite Fe3O4. The main diffraction peaks at 2θ° = 62.47, 56.97, 43.07, 35.45, 30.12 are related to the reflection planes of cubic spinel crystal structure of Fe3O4 at (440), (511), (400), (311), (220) respectively [43-50]. Fig. 2. FT-IR spectra of (a) GO, a e3O4@A/A-GO Fig. 3. XRD patterns of (a) GO, (b) Fe3O4@A/A-GO Fig. 4. XRD pat er ) , (b) Fe3O4@ /A-GO The SEM image of GO shows that graphene oxide nano sheets consists of randomly accumulated and wrinkled thin sheets. Also, the SEM image of Fe3O4@A/A-GO shows in Fig. 5(b) that the average diameter of Fe3O4 nanoparticles was about 35.43 nm and indicate a regularly spherical morphology on A/A-GO [51-52]. Fig. 5. SEM images of (a) GO, (b) Fe3O4@A/A-GO The Magnetic behavior of the Fe3O4@A/A-GO was recorded using a VSM. As shown in the Fig. 6a, the saturation magnetization of Fe3O4@A/A-GO was found to be 42 emu g. This amount of a saturation magnetization value, the magnetized nanocomposite is expected to have considerable paramagnetism to make it magnetically separable from reaction mixture [32, 48]. Nickel sep ration in blood by Fe3O4@A/A-GO A ne T égouë , et al 10 Anal. Method Environ. Chem. J. 3 (1) (2020) 5-16 2θ° = 62.47, 56.97, 43.07, 35.45, 30.12 are related to the reflection planes of cubic spinel crystal structure of Fe3O4 at (440), (511), (400), (311), (220) respectively [43-50]. The SEM image of GO shows that graphene oxide nano sheets consists of randomly accumulated and wrinkled thin sheets (Fig. 4a). Also, the SEM image of Fe3O4@A/A-GO shows in Figure 4(b) that the average diameter of Fe3O4 nanoparticles was about 35.43 nm and indicate a regularly spherical morphology on A/A-GO [51-52]. The Magnetic behavior of the Fe3O4@A/A-GO was recorded using a VSM. As shown in the Figure 5, the saturation magnetization of Fe3O4@A/AGO was found to be 42 emu g-1. This amount of a saturation magnetization value, the magnetized nanocomposite is expected to have considerable paramagnetism to make it magnetically separable from reaction mixture [32, 48]. 3.2. Optimization of extraction procedure For efficient extraction of nickel ions in human blood samples, the main parameters must be studied. The effective features such as, pH, sample volume, amount of Fe3O4@A/A-GO, shacking time and interferences ions must be optimized. Chemical bonding was strongly depended on the pH solutions. By procedure, the effect of pH on extraction of nickel through the amine functional group of Fe3O4@A/A-GO was evaluated. For this purpose, the different pH values from 1 to 11 with nickel concentration of 0.2-7 μg L-1 as LLOQ and ULOQ was examined according to DS-μ-SPE procedure. Obviously, the maximum of extraction Fig. 4. XRD patterns of (a) GO, (b) Fe3O4@A/A-GO The SEM image of GO shows that graphene oxide nano sheets consists of randomly accumulated and wri kled thin sheets. Also, the SEM image of Fe3O4@A/A-GO shows in Fig. 5(b) that the average diameter of Fe3O4 nanoparticles was about 35.43 nm and indicate a regularly spherical morphology on A/A-GO [51-52]. Fig. 5. SEM images of (a) GO, (b) Fe3O4@A/A-GO The Magnetic behavior of the Fe3O4@A/A-GO was recorded using a VSM. As shown in the Fig. 6a, the saturation magnetization of Fe3O4@A/A-GO was found to be 42 emu g. This amount of a saturation magnetization value, the magnetized nanocomposite is expected to have considerable paramagnetism to make it magnetically separable from reaction mixture [32, 48]. Fig. 6. VSM curve of Fe3O4@A/A-GO 3.2. Optimization of extraction procedure For efficient extraction of nickel ions in human blood samples, the main parameters must be studied. The effective features such as, pH, sample volume, amount of Fe3O4@A/A-GO, shacking time and interferences ions must be optimized. Chemical bonding was strongly depended on the pH solutions. By procedure, the effect of pH on extraction of nickel through the amine functional group of Fe3O4@A/AGO was evaluated. For this purpose, the different pH values from 1 to 11 with nickel concentration of 0.2-7 μg L as LLOQ and ULOQ was examined according to DS-μ-SPE procedure. Obviously, the maximum of extraction efficiencies for Ni (II) ions based on Fe3O4@A/A-GO were obtained at pH range of 8.2, and then the recoveries were decreased by increasing or decreasing of pH (Fig. 7). In optimized conditions, the effect of sample volume of blood on nickel extraction was studied between 2-20 mL. The results showed, the optimum extraction was achieved for 12 mL of blood sample and 15 mL of standard samples, So, 10 mL of sample volume was used for further studies. Also, the amount of Fe3O4@A/A-GO for nickel extraction was evaluated by DS-μ-SPE procedure. Based on experimental results, 10 mg of Fe3O4@A/A-GO was selected as optimum point. The sonication time for extraction of Ni(II) was studied in optimized pH. Based on previous research, kind and size of adsorbents are the most important factors for extraction and sonication time. Therefore, the effect of sonication in blood samples was examined by DS-μ-SPE procedure. The results showed, the maximum efficiency was achieved at 5 min. The concentration of Interfering ions such as Na, K, Mg, Ca, Cu, Zn, Co, Al, Hg, SO3, I, NO3-, Cl and F caused less than ± 5% deviation in the recovery of Ni(II) as the tolerance limit. So, the interfering ions has no effect on the recovery efficiencies of Ni(II) in blood samples. Fig. 5. VSM curve of Fe3O4@A/A-GO Fig. 4. SEM images of (a) GO, (b) 3 4 A/A-GO 11 Nickel separation in blood by Fe3O4@A/A-GO Anne Trégouët, et al efficiencies for Ni (II) ions based on Fe3O4@A/AGO were obtained at pH range of 8.2, and then the recoveries were decreased by increasing or decreasing of pH (Fig. 6). In optimized conditions, the effect of sample volume of blood on nickel extraction was studied between 2-20 mL. The results showed, the optimum extraction was achieved for 12 mL of blood sample and 15 mL of standard samples, So, 10 mL of sample volume was used for further studies. Also, the amount of Fe3O4@A/AGO for nickel extraction was evaluated by DS-μSPE procedure. Based on experimental results, 10 mg of Fe3O4@A/A-GO was selected as optimum point. The sonication time for extraction of Ni(II) was studied in optimized pH. Based on previous research, kind and size of adsorbents are the most important factors for extraction and sonication time. Therefore, the effect of sonication in blood samples was examined by DS-μ-SPE procedure. The results showed, the maximum efficiency was achieved at 5 min. The concentration of Interfering ions such as Na+, K+, Mg2+, Ca2+, Cu2+, Zn2+, Co2+, Al3+, Hg2+, SO3 2-, I-, NO3 -, Cland Fcaused less than ± 5% deviation in the recovery of Ni(II) as the tolerance limit. So, the interfering ions has no effect on the recovery efficiencies of Ni(II) in blood samples. 3.3 . Validation methodology Many methods was used for validation of methodology by SPE [53-55]. The analytical results of the developed DS-μ-SPE procedure were shown method at optimum conditions (Table 1). After sample preparation, Ni concentration in human blood and standard samples was determined by ET-AAS. The Human blood, serum and plasma as a real sample was used for determination of Ni by DS-μ-SPE procedure. The results was verified by analyzing the spiked samples with standard concentration of Ni (II) in human samples (Table 2). Based on results, a high and favorite recovery was obtained by spiking samples which confirms the accuracy of results in difficulty matrix. The Fig. 6. The effect of pH on nickel extraction by DS-μ-SPE procedure Fig.7. The effect of pH on nickel extraction by S-μ-SPE procedure 3.3 . Validation methodology Many methods was used for validation of methodology by SPE [ 53-55]. The analytical results of the developed DS-μ-SPE procedure were shown method at optimum conditions (Table 1). After sample preparation, Ni concentration in human blood and standard samples was determined by ET-AAS. The Human blood, serum and plasma as a real sample was used for determination of Ni by DS-μ-SPE procedure. The results was verified by analyzing the spiked samples with standard concentration of Ni (II) in human samples (Tables 2). Based on results, a high and favorite recovery was obtained by spiking samples which confirms the accuracy of results in difficulty matrix. The recoveries of spiked samples demonstrated that the results was satisfactory for Ni analysis by DS-μ-SPE. In order to validate the method, the extraction efficiency for intra-day and inter day analysis was evaluated by spiking samples (Table 3). Table 1. Analytical results for Ni(II) extraction based on Fe3O4@A/A-GO by DS-μ-SPE Element SV LR R LOD (n = 10) RSD b (%) EF Ni(II) 10 0.157.2 0.9997 0.035 2.8% 19.8 a sample volume (mL), b Linear rang (μμg L) , Limit of detection (μμg L) , d Relative standard deviation, F enrichment factor(EF), 0 20 40 60 80 100 120 2 3 4 5 6 7 7.5 8 8.5 9 10 Re co ve ry (% )


Introduction
Nickel (Ni) is one of the toxic compounds in water contamination and caused to acute and chorionic effect in human bodies [1]. Ni(II) are released into environment from waste of different chemical factories such as battery manufacturing, mining and electroplating. The air of near factories contain the significant amount of heavy metals such as nickel and caused to adverse effects in environment 6 Anal. Method Environ. Chem. J. 3 (1) (2020) [5][6][7][8][9][10][11][12][13][14][15][16] Toxic effects of oral exposure to nickel usually involve the kidneys with some evidence from animal studies showing a possible developmental/ reproductive toxicity effect (ATSDR). Normal range for Ni in healthy peoples is 0.2 µgL -1 in serum and less than 3.0 µgL -1 in human urine. A national health and nutritional examination survey (NHNES) of hair found mean nickel levels of 0.39 mg L -1 , with 10% of the population having levels less than1.5 mg L -1 [4,5]. The vary techniques was used for measurement of nickel (Ni 2+) in human bodies such as serum and blood samples. Due to previous studies, the flame atomic absorption spectrometry (F-AAS) and electrothermal atomic absorption spectrometry (ET-AAS) were reported for analysis of heavy metals such nickel (Ni) which was suitable for the determination in biological matrixes [6][7][8][9] . Occasionally, the atom trapping flame atomic absorption spectrometry (AT-FAAS) and fluorescence spectrometry (XRF) also was used for heavy metal determaintion such as nickel in biological samples [10,11]. Also, different methods such as, polarography [12], inductively coupled plasma-optical emission spectrometry (ICP OES, inductively coupled plasma-mass spectrometry (ICP-MS) [13,14], and spectrophotometry [15] were used for nickel analysis in human samples [16]. But as difficulty matrixes in human biological samples such as blood or serum, the sample preparation is required to separation nickel ions from liquid phases. The different sample preparation technology exist for nickel extraction from blood samples such as liquidliquid extraction (LLE) [17], solid-phase extraction (SPE) and functionalized magnetic-SPE [18], dispersive solvent by liquid-liquid microextraction method (DLLME) [19], phase extraction based on cloud point (CPE) [20], dispersive solid phase microextraction (D-SPME) [21], ultrasoundassisted solid phase extraction (USA-SPE) [22]. Recently, the dispersive sorbent in the liquid phase was presented as micro SPE (D-μ-SPE) for separation/determination of nickel in water and biological samples [23]. Some advantages such as high extraction efficiency, simple usability, fast and low time caused to select as a favorite technique for metal extraction. The sorbent characterizations are a main factor which was effected on heavy metal extraction by the D-μ-SPE procedure. The selective of favorite nanosturactures improved recovery. The different nanosorbent such as, carbon nanotubes (CNTs), graphene / graphene oxide sheets (NG, NGO) and silica (MSN) were used for extraction of Ni in blood samples [24][25][26]. Between them, the NGO was reported as efficient sorbents for metal extraction due to their surface properties. In this study, a novel sorbent based on Fe 3 O 4supported amine/amide-functionalized graphene oxide (Fe 3 O 4 @A/A-GO) was used for separation/ speciation Ni(II) from human blood samples by dispersive sonication micro solid phase extraction (DS-μ-SPE) at pH=8.0. The method was developed in blood samples by ET-AAS.

Instrumental
The graphite furnace atomic absorption spectrophotometer (GF-AAS, 932 GBC, Aus) was used for nickel determination in blood samples with the Avanta software. The linear range of 3.0-150 µg L -1 (peak Area, Abs=1.91) was selected for Ni in optimized light. The current and wavelength of HCL lamp was adjusted 3.0 mA and 228.8 nm, respectively. The auto-sampler of spectrophotometer (Pal 3000) was used for micro injection (1µL) of sample volumes to furnace tube after adjusted injector. The pH of the sample was digitally calculated by Metrohm pH meter (Swiss). Fourier transform infrared (FT-IR) spectra were recorded from KBr pellets using a Perkin Elmer Spectrum 65 FT-IR spectrophotometer. Powder X-ray diffraction (XRD) was conducted on a Panalytical X'Pert PRO X-ray diffractometer. Scanning electron microscopy (SEM) images were obtained using a Tescan Mira-3 Field Emission Gun Scanning Electron Microscope (SEM). Magnetic

Synthesis of graphene oxide (GO)
Graphite black powder was oxidized to GO following the modified Hummers method in the several steps [27][28][29]. Fore pre oxidation of graphite powder, 250 mL of H 2 SO 4 was added to 5 g of graphite powder and the resulting mixture was stirred for 24 h. After 24 h, 30 g of KMnO 4 was added to the mixture stirring for 72 h at 50 °C. Next, a solution of 45 mL of H 2 O 2 (30%) in 400 mL of deionized water was added to the mixture after which the brown color of the mixture turned into bright yellow. The GO dark solution was centrifuged and washed with deionized water and 10% HCl solution, and dried at 65 °C .

Preparation of nanomagnetic Fe 3 O 4supported Amine/Amide-functionalized graphene oxide (Fe 3 O 4 @A/A-GO) 2.4.1. Acyl-chlorination of graphene oxide
In the first step, thionyl chloride 60 mL was added GO (0.3 g) and stirred in at 70°C for 24 h. After acyl-chlorination reaction, the Acyl-chlorination of GO was washed with THF four times and the precipitated dried at 65°C [30].

Functionalization of graphene oxide with Ammonia
A total of 1 g of Ammonia was added to Acylchlorinated GO (0.3 g). Then, the mixture was refluxed for 72 h at 120°C under argon condition. Finally, the mixture reaction washed with DI water. A dark powder of Amine/Amide-functionalized graphene oxide was obtained [31].

Nano magnetization of Amine/Amide-funcftionalized graphene oxide
The Fe 3 O 4 -supported Amine/Amide-functionalized graphene oxide (Fe 3 O 4 @A/A-GO) nanoparticles were synthesized by co-precipitation of FeCl 2 ·4H 2 O and FeCl 3 ·6H 2 O, in the presence of Amine/Amide-grafted graphene oxide. First, a solution of FeCl 2 ·4H 2 O and FeCl 3 ·6H 2 O was prepared with a molar ratio of 2:1. The weight ratio of GO to FeCl 3 in the nano composite was 1 per 20 (mGO: mFeCl 3 = 1:20). Afterward, 10 mg of Amine/Amide-grafted graphene oxide in 15 mL of DI water was ultrasonicated for 20 min. 12.5 mL solution of FeCl 2 ·4H 2 O (125 mg) and FeCl 3 ·6H 2 O (200 mg) in deionized water (10 mL) was added to the reaction mixture at room temperature. In order to raise pH value to 11, an aqua solution of 30% ammonia was added in the reaction mixture at 70 °C . Then, the reaction mixture was allowed to cool to room temperature and Fe 3 O 4 @A/A-GO washed five times with DI H 2 O dried at 70 °C [32].

Characterization of the nano Fe 3 O 4 @A/A-GO
The Fe 3 O 4 -supported Amine/Amide-functionalized graphene oxide nanomagnetic (Fe 3 O 4 @A/A-GO) was synthesized according to the synthetic route shown in Figure 1. Firstly, graphite was oxidized to GO by the modified Hummers method [27][28][29]. Afterward, the GO was amination and amidation with Ammonia by according to our previously reported procedure [30,31]. Finally, the resulted Amine/Amide-functionalized graphene oxide (Fe 3 O 4 @A/A-GO) was nano-magnetized by coprecipitation of ferrous (Fe 2+ ) and ferric (Fe 3+ ) ions in the presence of A/A-GO to afford the target adsorbent Fe 3 O 4 @A/A-GO [32]. Figure 2 shows the FT-IR spectra of GO and Fe 3 O 4 @A/A-GO. The broad peak in the range between 2600-3500 cm -1 in the IR spectra of these compounds is related to the (O-H stretching)  Figure 3 (a,b). GO has the two main peak at 2θ = 11.5°, 42.58 o are related to the diffraction planes of (002) and (100) respectively, which can be observed in the XRD patterns of both GO and Fe 3 O 4 @A/A-GO [40][41][42]. As shown in Figure 3  The main diffraction peaks at 3500 cm -1 in the IR spectra of these compounds is related to the (O-H stretching) vibration of carboxylic and enolic functionalities [35]. The peaks at 3420, 1719, 1621, and 1060 cm -1 shown in the spectra of GO and Fe 3 O 4 @A/A-GO are ascribed to the (C-O stretching), (C=C stretching), (C=O stretching), and (O-H stretching) respectively [33,34]. Also, The absorption bands at 3360, 3181, 1630, and 1234 cm-1 shown in the spectra of GO and amination GO are ascribed to the stretching bands ν(O-H), ν(N-H), ν(C=O), and ν(C-N) respectively. In the spectrum of Fe 3 O 4 @A/A-GO, the peaks observed at around 628 and 583 cm -1 are related to the Fe-O stretching vibration [37][38][39]. These results prove that the successful Amination and amidation of GO and synthesis of Fe 3 O 4 @A/A-GO.

Fig. 2. FT-IR spectra of (a) GO, and (b) Fe 3 O 4 @A/A-GO
The XRD patterns of GO, and Fe 3 O 4 @A/A-GO are demonstrated in Fig. 4. GO has the two main peak at 2θ = 11.5°, 42.58 o are related to the diffraction planes of (002) and (100) respectively, which can be observed in the XRD patterns of both GO and Fe 3 O 4 @A/A-GO [40][41][42]. As shown in Fig. 4 (b), the peaks at 2θ = 24° and 42.58 o It is evident from the XRD pattern of Fe 3 O 4 @A/A-GO that was proved the presence of GO [42]. The main peaks at 2θ° = 30.  The SEM image of GO shows that graphene oxide nano sheets consists of randomly accumulated and wrinkled thin sheets. Also, the SEM image of Fe 3 O 4 @A/A-GO shows in Fig. 5(b) [43][44][45][46][47][48][49][50].
The SEM image of GO shows that graphene oxide nano sheets consists of randomly accumulated and wrinkled thin sheets (Fig. 4a). Also, the SEM image of Fe 3 O 4 @A/A-GO shows in Figure 4(b) that the average diameter of Fe 3 O 4 nanoparticles was about 35.43 nm and indicate a regularly spherical morphology on A/A-GO [51][52].
The Magnetic behavior of the Fe 3 O 4 @A/A-GO was recorded using a VSM. As shown in the Figure  5, the saturation magnetization of Fe 3 O 4 @A/A-GO was found to be 42 emu g -1 . This amount of a saturation magnetization value, the magnetized nanocomposite is expected to have considerable paramagnetism to make it magnetically separable from reaction mixture [32,48].

Optimization of extraction procedure
For efficient extraction of nickel ions in human blood samples, the main parameters must be studied. The effective features such as, pH, sample volume, amount of Fe 3 O 4 @A/A-GO, shacking time and interferences ions must be optimized. Chemical bonding was strongly depended on the pH solutions. By procedure, the effect of pH on extraction of nickel through the amine functional group of Fe 3 O 4 @A/A-GO was evaluated. For this purpose, the different pH values from 1 to 11 with nickel concentration of 0.2-7 μg L -1 as LLOQ and ULOQ was examined according to DS-μ-SPE procedure. Obviously, the maximum of extraction  The Magnetic behavior of the Fe 3 O 4 @A/A-GO was recorded using a VSM. As shown in the Fig. 6a, the saturation magnetization of Fe 3 O 4 @A/A-GO was found to be 42 emu g -1 . This amount of a saturation magnetization value, the magnetized nanocomposite is expected to have considerable paramagnetism to make it magnetically separable from reaction mixture [32,48].

Optimization of extraction procedure
For efficient extraction of nickel ions in human blood samples, the main parameters must be studied. The effective features such as, pH, sample volume, amount of Fe 3 O 4 @A/A-GO, shacking time and interferences ions must be optimized. Chemical bonding was strongly depended on the pH solutions. By Therefore, the effect of sonication in blood samples was examined by DS-μ-SPE procedure. The results showed, the maximum efficiency was achieved at 5 min. The concentration of Interfering ions such as Na + , K + , Mg 2+ , Ca 2+ , Cu 2+ , Zn 2+ , Co 2+ , Al 3+ , Hg 2+ , SO 3 2-, I -, NO 3 -, Cland Fcaused less than ± 5% deviation in the recovery of Ni(II) as the tolerance limit. So, the interfering ions has no effect on the recovery efficiencies of Ni(II) in blood samples.

. Validation methodology
Many methods was used for validation of methodology by SPE [53][54][55]. The analytical results of the developed DS-μ-SPE procedure were shown method at optimum conditions (Table  1). After sample preparation, Ni concentration in human blood and standard samples was determined by ET-AAS. The Human blood, serum and plasma as a real sample was used for determination of Ni by DS-μ-SPE procedure. The results was verified by analyzing the spiked samples with standard concentration of Ni (II) in human samples ( Table  2). Based on results, a high and favorite recovery was obtained by spiking samples which confirms the accuracy of results in difficulty matrix. The

. Validation methodology
Many methods was used for validation of methodology by SPE [53][54][55]. The analytical results of the developed DS-μ-SPE procedure were shown method at optimum conditions (Table 1). After sample preparation, Ni concentration in human blood and standard samples was determined by ET-AAS. The Human blood, serum and plasma as a real sample was used for determination of Ni by DS-μ-SPE procedure. The results was verified by analyzing the spiked samples with standard concentration of Ni (II) in human samples (Tables 2). Based on results, a high and favorite recovery was obtained by spiking samples which confirms the accuracy of results in difficulty matrix. The recoveries of spiked samples demonstrated that the results was satisfactory for Ni analysis by DS-μ-SPE. In order to validate the method, the extraction efficiency for intra-day and inter day analysis was evaluated by spiking samples (Table 3). 19.8 a sample volume (mL), b Linear rang ( g L −1 ) , C Limit of detection ( g L −1 ) , d Relative standard deviation, F enrichment factor(EF), recoveries of spiked samples demonstrated that the results was satisfactory for Ni analysis by DS-μ-SPE. In order to validate the method, the extraction efficiency for intra-day and inter day analysis was evaluated by spiking samples (Table 3).

Conclusions
A fast and efficient method based on Fe 3 O 4 @A/A-GO as adsorbent was used for preconcentration, separation of trace Ni (II) in human blood samples by DS-μ-SPE procedure. The mechanism of extraction was achieved by interaction between negative charge (-) of amine group of Fe 3 O 4 @A/ A-GO with positive charge (+) of nickel ions in favorite pH (Ni 2+ ….. NH 2 ). After extraction, the concentration of nickel was determined by ET-AAS technique. Finally, the developed method has low ion interference, simple usage with low LOD, favorite RSD(%) values and good LR with high recoveries for Ni extraction in blood samples (>95%). Therefore, the proposed method can be considered as applied techniques for Ni separation and determination in blood samples by DS-μ-SPE coupled to ET-AAS.

Acknowledgements
The authors wish to thank the Research Institute of Petroleum Industry (RIPI) for financial support to