Magnetic Nanoparticles Coated with Ionic Liquid as a Sorbent for Solid Phase Extraction of Chromium(VI) Prior to Its Determination by Electrothermal Atomic Absorption Spectrometry

A simple, sensitive and reliable method for the separation, preconcentration and determination of ultra trace amounts of chromium species has been developed. Chromium(VI) in aqueous sample was reacted with 9-phenyl-2,3,7-trihydroxy-6-fluorone to produce a chelate at pH of 5.0 and extracted onto the magnetic nanoparticles coated with the ionic liquid 1-hexadecyl-3- methylimidazolium bromide. The trapped analyte was back extracted using 350 µL of nitric acid solution (2 mol L-1) and was determined by electrothermal atomic absorption spectrometry. Total chromium was determined by oxidizing CrIII to CrVI using KMnO4 in acidic media. Under the optimum conditions, the method exhibited a linear dynamic range of 0.01-0.50 µg L-1 with an enhancement factor of 112 and a detection limit of 0.003 µg L-1 for CrVI. The coefficient of variation (n = 6) at 0.3 µg L-1 concentration level of CrVI was 3.2%.


Introduction
The toxicity, reactivity and biological properties of chromium depend on its chemical form.Chromium is found in two most stable oxidation states of Cr III and Cr VI .2][3][4] The concentration of Cr III and Cr VI in human serum of a normal healthy person is in the range of 0.52-0.66and 0.22-0.82µg L -1 respectively, 1 but the level of chromium in the serum of diabetic person is lower.Thus, from medical point of view, accurate and precise determination of chromium in biological sample including human serum is important.Chromium is also known as a major water pollutant. 5The United States Environmental Protection Agency (USEPA) has set a total concentration of 0.1 mg L -1 chromium in drinking water as the maximum allowable contaminant level whereas World Health Organization (WHO) states the guideline values of 50 µg L -1 Cr VI . 6Consequently, the development of an accurate, easy and sensitive method for the determination of chromium species in environmental waters and biological samples is an attractive task in analytical chemistry.
Various analytical techniques such as atomic absorption spectrometry (AAS), [7][8][9] spectrophotometry, 10 inductively coupled plasma mass spectrometry (ICP-MS), [11][12][13] X-ray fluorescence (XRF) spectrometry 14 and inductively coupled plasma optical emission spectrometry (ICP OES) 15 have been used for the determination of chromium in different samples.However, these methods only measure the total chromium amount.Thus, due to the presence of low level of chromium in real samples, complexity of the matrices and the importance of measurement of chromium species, a separation and preconcentration step prior to its determination is often a necessary step.Procedures reported in the literature for the speciation of chromium are generally liquid-liquid extraction (LLE), 16 solid phase extraction (SPE), [17][18][19] cloud point extraction (CPE), 1,20 hollow fiber liquid phase microextraction (HF-LPME), 21 dispersive liquid-liquid microextraction (DLLME), [22][23][24] solidified floating organic drop microextraction (SFODME), 25 capillary electrophoresis 26 and electrochemical methods. 279][30][31] However, the nature of sorption material plays a unique role in SPE process because it determines the analytical sensitivity, affinity, capacity and repeatability of the method.The most important sorbents used in the SPE procedures for speciation of chromium are Ambersorb 563 resin, 17 activated carbon, 32 cellulose 5 and TiO 2 . 33In SPE methods, the extraction is often performed either in batch or in column modes.However, when the extraction is done in the batch mode, separation of traditional sorbent from the sample is difficult and time consuming, whereas, in column mode, the flow rate of sample through the column is limited and thus, for obtaining a high preconcentration factor (PF) the passage of large volume of sample through the column is tedious.5][36] Magnetic nanoparticles (MNPs) with the large and high surface area make it possible to obtain high PF from smaller sample volume.However, the basic disadvantage of bare MNPs is the lack of the selectivity for target analyte.To overcome this problem, magnetic nanoparticles can be modified with different organic ligands, 37 imprinted polymers 38 and surfactants. 39ecently, ionic liquids (ILs) have attracted the analytical chemists as a new class of coating materials in SPE. 40,41ILs are organic salts of organic cations paired with inorganic or organic anions with melting points often less than 100 °C.2,43 However, according to our literature survey the use of ILs for coating the nanoparticles are still in its infancy state.In 2011, Absalan et al. 44 coated MNPs with IL (1-hexyl-3methylimidazolium bromide) and use it as the sorbent for removal of reactive red-120 and 4-(2-pyridylazo)resorcinol from aqueous samples.In 2012, Farahani et al. 30 deposited hydrophobic IL (1-hexyl-3-methylimidazoliumhexafluorop hosphate) on the surface of magnetic nanoparticles and used it for SPE of lead and cadmium.In 2013, Amjadi et al. 29 used modified IL-coated TiO 2 nanoparticles as a new solid phase extraction sorbent for preconcentration of trace nickel.In this study, the sorption/desorption possibility of Cr VI -9-phenyl-2,3,7-trihydroxy-6-fluorone complex onto the MNPs coated with 1-hexadecyl-3-methylimidazolium bromide (C 16 mimBr) IL was considered and a rapid SPE method coupled with electrothermal atomic absorption spectrometry (ETAAS) was developed for separation/ preconcentration and speciation of chromium species.Total chromium was determined after oxidation of Cr III to Cr VI and the amount of Cr III was determined from the difference of concentration of total chromium and Cr VI .

Experimental
Apparatus A Varian model SpectrAA 220Z (Mulgrave, VIC, Australia) Zeeman atomic absorption spectrometer equipped with an auto sampler (PSD 120) and a graphite tube atomizer (GTA 120) was used for all the measurements throughout this study.A computer was used to record the absorbance signal profile.A Varian SpectrAA hollow cathode lamp for chromium was operated at 357.9 nm with 7 mA current and a spectral bandwidth of 0.5 nm.The furnace tube was a standard plateau tube with a pyrolytic graphite coating.The Zeeman background correction was used for all measurements.The furnace temperature program applied was as recommended by the manufacturer (Table 1).Peak height measurement was used for all the quantifications.The sample injection volume of 15 µL along with 10 µL modifier was used in all experiments.The pH measurements were carried out with a Metrohm pH meter (model 691, Herisau, Switzerland) using a combined glass calomel electrode.In addition, for magnetic separations a strong neodymium-iron-boron (Nd 2 Fe 12 B) magnet (1.31 T) was used.

Standard solution and reagents
Iron(II) chloride tetrahydrate, iron(III) chloride hexahydrate, and 1-hexadecyl-3-methylimidazolium bromide (C 16 mimBr) were purchased from Sigma-Aldrich (St. Louis, USA).Double distilled deionized water was used throughout this study.All glassware was washed with 10% (v/v) nitric acid and then rinsed with water before use.Stock standard solutions (1000 mg L -1 ) of Cr VI and Cr III were prepared by dissolving proper amounts of K 2 Cr 2 O 7 and CrCl 3 .6H 2 O in distilled water.Working solutions were prepared daily from the stock standard solutions by serial dilutions with double distilled water.9-Phenyl-2,3,7-trihydroxy-6-fluorone (phenylfluorone), ethanol, hydrochloric acid, nitric acid, potassium permanganate, sodium azide, acetonitrile, ammonia and all other chemical used were purchased from Merck Company (Darmstadt, Germany).Phenylfluorone solution in ethanol (1 × 10 -3 mol L -1 ) was prepared by dissolving 0.0320 g of phenylfluorone in ethanol containing 1 mL concentrated HCl and diluting to 100 mL upon addition of ethanol.A Pd/Mg modifier was prepared from the palladium modifier solution for ETAAS and Mg(NO 3 ) 2 .6H 2 O according to the literature. 45

Synthesis of Fe 3 O 4 nanoparticles
Fe 3 O 4 MNPs were prepared by the coprecipitation method. 35A 50 mL of an aqueous solution containing 5.2 g of FeCl 3 .6H 2 O and 2.0 g of FeCl 2 .4H 2 O was heated at 80 °C for 15 min.Then, 10 mL of concentrated NH 3 was added dropwise.N 2 gas was used as the protective gas in the whole experiment.After completion of the reaction, the black precipitate was collected by an external magnetic field, washed with water and ethanol and dried in oven at 80 °C.

Preparation of magnetite nanoparticles coated with ILs
Twenty five milliliter of C 16 mimBr IL (2.0 g L -1 ) was added to the Fe 3 O 4 MNPs (0.5 g) in a 50 mL beaker.The pH was adjusted to 10.0 with sodium hydroxide solution (1.0 mol L -1 ) and was mixed thoroughly by mechanical stirrer for 20 min.In this stage, the nanoparticles were suspended in the mixture and covered with the IL.Then, the modified MNPs were isolated by application of an external magnetic field, washed with water and dried at room temperature for 24 h.The sorbent was then used for the further studies.

Preparation of real samples
Water samples were filtered through 0.45 µm Millipore filter.Their pHs were adjusted to 5.0 and treated according to the extraction procedure given below.Frozen serum samples of the diabetic persons were provided from a hospital in Yazd.After reaching the ambient temperature, a small amount of acetonitrile was added to 5.0 mL of serum sample to precipitate the protein contents of it and centrifuged for 10.0 min at the rate of 5000 rpm.The supernatant was diluted with ultrapure water at a ratio 1:10 and was treated according to the given procedure.

Extraction procedure Determination of Cr VI
To 40 mL of sample or standard solution containing of Cr VI and Cr III , 1.0 mL of a 1.0 × 10 -3 mol L -1 phenylfluorone was added and the pH was adjusted to 5.0 upon addition of 2 mL acetate buffer.Then, 15 mg modified MNPs was added and the mixture was stirred thoroughly for 15.0 min.At this stage, phenylfluorone-Cr VI chelate was adsorbed onto the modified MNPs.Subsequently, a strong magnet was placed at the bottom of the beaker and the sorbent containing the analyte was trapped.The bulk aqueous phase was easily decanted and the sorbed analyte was desorbed by addition of 350 µL of nitric acid (2.0 mol L -1 ) and stirred for 2 min.Finally, the sorbent was retained with the help of a magnet and the supernatant solution was transferred into the electrothermal cup.Then, 15 µL of it along with 10 µL of the Pd/Mg modifier was injected into the graphite tube of ETAAS for the quantification of analyte.

Determination of total chromium and Cr III
To determine the total concentration of chromium in the solution, Cr III was efficiently oxidized to Cr VI according to the given procedure; 4 i.e., 4 to 5 drops of KMnO 4 solution (0.02 mol L -1 ) and an adequate amount of sulfuric acid to make its concentration 0.1 mol L -1 were added to 40 mL of sample, the beaker was covered with a watch glass and heated at 45 °C for 15 min to completely oxidize the Cr III to Cr VI .After the solution was cooled, the excess of KMnO 4 was reduced upon drop wise addition of sodium azide solution (2.0%, m/v) until the pink color of the solution was removed.The solution was then treated according to the given procedure in determination of Cr VI .The concentration of Cr III was determined from the difference of concentration of total chromium and Cr VI .

Results and Discussion
9-Phenyl-2,3,7-trihydroxy-6-fluorone (phenylfluorone) reacts with chromium(VI) and produces a stable and sensitive 2:1 purplish red chelate at pH 4.7-6.6 (Figure 1). 46,47Phenylfluorone had been used for direct spectrophotometric determination 47 and cloud point extraction of chromium(VI). 46,48In the preliminary work, it was established that the chelate of Cr VI with phenylfluorone can be extracted onto the IL coated magnetic nanoparticles, while Cr III remains in aqueous phase.Then, the sorbent was characterized and an extraction system was designed.In order to obtain a high enrichment factor and appropriate situation for the speciation of chromium, different parameters affecting the chelate formation, extraction and analysis process were optimized in a univariable approach.

Characterization of the sorbent
The Fourier transform infrared spectroscopy (FTIR) spectra of C 16 mimBr IL, MNPs and MNPs modified with IL were recorded by KBr pellet method (Figure 2).The spectra of IL shows peak at 1168, 1467 and 1575 cm -1 corresponding to C−C, C=C and C=N stretching vibrations, respectively.The same peaks with slight shift to lower wavenumbers (i.e., to 1167, 1444 and 1571 cm -1 , respectively) are observed in modified nanoparticles, indicating that the IL is coated on the magnetic nanoparticles.
The surface morphology of the magnetic nanoparticles and nanoparticles modified with IL were characterized by scanning electron microscopy (SEM) (Figure 3).As it is demonstrated the size of magnetic nanoparticles after modification with IL is not significantly changed and is still in dimension of nanometers., the effect of sample pH on the extraction of Cr VI was investigated.The results (Figure 4) demonstrated that the extraction recovery of Cr VI is maximized and constant in the pH range 4.0-7.5 which corresponds to the region where the CrO 4 2− species is predominate.The decrease in the extraction of Cr VI at pH higher than 7.5 can be related to the decomposition of Cr VI -phenylfluorone chelate and instability of complex at alkaline media.The negligible extraction of Cr III in all studied pHs is because the phenylfluorone is a selective ligand for Cr VI and does not form a stable chelate with Cr III .Thus, Cr VI can be extracted as Cr VI -phenylfluorone chelate while Cr III remains in the aqueous phase.This suggests that it is possible to separate the Cr VI from Cr III in the whole pH range and the extraction of Cr VI reaches its maximum in the pH range of 4.0-7.5.Therefore, the pH of 5.0 was selected for the subsequent works.
The effect of the concentration of chelating agent on the extraction recovery of Cr VI was investigated and   It is well known that the convection induced by stirring of the sample causes fast mass transfer to occur.Thus, a fast equilibrium between the sample solution and modified MNPs can be attained by proper stirring of the solution.The effect of stirring rate was investigated in the range of 300-1200 rpm.It was observed that, 1000 rpm was sufficient to achieve a quantitative extraction of the analyte in 15.0 min.
The type, concentration and volume of the desorbing solution have a considerable effect on the extraction efficiency.The eluent must be capable of complete desorption of analyte in minimum volume in a short time and it must not interfere in the measurement of analyte.The possibility of desorption of chromium complex by 500 µL of different eluent including nitric acid, hydrochloric acid and acetic acid all at a concentration of 2.0 mol L -1 was considered.It was found that nitric acid has the maximum capability in desorbing the analyte, whereas, the recoveries with hydrochloric acid and acetic acid were about 60% and 40% of that of nitric acid, respectively.The effect of nitric acid concentration in the range of 0.5-5.0mol L -1 on the recovery was then studied.The results implied that the extraction recovery increased with an increase in nitric acid concentration up to 2.0 mol L -1 and then remained constant at higher concentrations.Therefore, nitric acid with a concentration of 2.0 mol L -1 was chosen as the desorbing solution for the subsequent studies.Furthermore, the influence of the volume of nitric acid on the recovery of Cr VI was studied by varying its volume in the range of 100-500 µL.The results demonstrated that chromium was quantitatively desorbed from modified MNPs in volumes of 350 µL and larger.For achieving high enrichment factor, the smaller volume of eluent in which the recovery was quantitative (350 µL) was chosen.
The effect of desorption time was also evaluated by varying desorption time between 0.5 and 10.0 min.It was found that desorption of analyte from the sorbent is relatively fast and 2.0 min stirring was sufficient for the quantitative recovery of the analyte.
The recovery values for chromium were dependent on the mass of the sorbent.The recovery increased by an increase in sorbent mass up to 10 mg and then remained constant and quantitative by further increase in amount of the sorbent.A mass of 15 mg of the sorbent was selected as the optimum amount for the further studies.
The effect of ionic strength on the extraction (or stripping or removing) of chromium complex from the modified MNPs sorbent was investigated by varying the NaCl concentration between 0.0-1.0mol L -1 .The results proved that the extraction efficiency is independent of the salt concentration.Thus, the method can be applied for quantitative separation and preconcentration of chromium from saline solutions.Further studies were done without salt addition.
Demonstrating the capability of the extraction system in obtaining high preconcentration factor is an important  aspect of method development as it shows the possibilities of recovery of analyte from a large sample volume.An increase in the ratio of the volume of the aqueous phase to the eluent will increase the preconcentration factor but it may reduce the extraction efficiency.In order to study the effect of the sample volume on the extraction efficiency, 12 ng of Cr VI was extracted from different volumes of solution (10-60 mL) under constant the other experiment conditions.The results showed that the recovery was quantitative up to 40 mL of the sample and then decreased with further increase in sample volume.Thus, based on the volume of desorbing solution (350 µL) and the maximum sample volume that the extraction was quantitative (40 mL) a preconcentration factor of 114 was determined.

Sorbent capacity
The pH of sample solution (20 mg L -1 Cr VI , 40 mL) was adjusted to 5.0.2.0 mL of phenylfluorone and 50 mg of modified MNPs were added.Then, the mixture was stirred for 120 min and after collection of the sorbent by applying an external magnetic field the amount of chromium on the supernatant solution was determined by flame atomic absorption spectroscopy.The amount of chromium retained by the sorbent was determined from the difference of the concentration of chromium in the initial and final solutions.The capacity of the sorbent was found to be 11.5 mg g -1 .

Effect of foreign ions
The selectivity of developed SPE method for extraction and determination of chromium was demonstrated by studying the possibility of the effect of common interfering ions usually present in water and biological samples.For this purpose, 40 mL of the solution of 0.30 mg L -1 Cr VI and various amounts of interfering ions was preconcentrated and analyzed according to the recommended procedure.The tolerance limit of the coexisting ions was defined as the largest amount that make a variation of less than 5% in the recovery of the analyte.The results of these studies (Table 2) indicate that with the exception of Cu II , Al III , Ni II and Pb II ions, which interfere at the mole ratio of 50, the other ions at the given level show no significant interference in the determination of chromium.Furthermore, as EDTA forms ionic hydrophilic complexes with the interfering ions but does not have any affinity for Cr VI the tolerance limit of Cu II , Al III , Ni II and Pb II ions was improved to the mole ratio of 600 upon addition of EDTA.Thus, when the concentration of Cu II , Al III , Ni II and Pb II ions in the sample is more than 50 times of the analyte, the EDTA at a concentration of 2.0 × 10 -4 mol L -1 should be added to the sample prior to addition of sorbent.These results indicate that the developed method is selective for the determination of chromium(VI) at optimum conditions.

Analytical figures of merit
In the optimum conditions, a calibration graph was constructed for Cr VI by preconcentrating of several standard solutions according to the recommended procedure.The linear concentration range was found to be 0.01-0.50µg L -1 with a correlation coefficient of 0.9990.The equation of calibration graph was A = 1.622C + 0.036 (where A is the absorbance and C is the concentration of Cr VI in µg L -1 ).The limits of detection (LOD) and quantification (LOQ) defined as 3 S b /m and 10 S b /m (where S b is the standard deviation of the blank and m is the slope of the calibration graph) were 0.003 and 0.01 µg L -1 , respectively.The precision of proposed method was evaluated by subjecting a series of six solutions containing 0.3 µg L -1 Cr VI to the extraction and measurement process at same day.The coefficient of variation (CV, %) was found to be 3.2%.The enhancement factor defined as the slope ratio of two calibration curves with and without preconcentration was found to be 112.The closeness of enhancement and preconcentration factor to each other indicate that the extraction is quantitative (about 98% completed).

Application
To check the reliability of the proposed method for speciation of chromium, the method was applied to the determination of Cr VI and Cr III in several categories of water samples including river water (Zayandeh Rood River, Esfahan, Iran), subterranean water (the two-story subterranean of Ardestan, Esfahan, Iran), sea water and human serum.The accuracy of the method was examined by spiking the samples with two or three different levels of Cr III and Cr VI and calculating the recovery.The results of this investigation are given in Table 3.As it can be seen, the recoveries of added chromium species are good.Thus, the procedure is reliable and accurate for the analysis of chromium species at trace levels in water samples.Furthermore, the procedure was also applied to determination of chromium in a certified reference river water sample SLRS-1 with chromium concentration 0.36 ± 0.03 µg L -1 .
The concentration of chromium in this sample was found to be 0.35 ± 0.02 µg L -1 .Thus at 95% confidence level there is no significant difference between the obtained value and the accepted one.Thus, the procedure is reliable for analysis of chromium in the sample types studied.

Comparison with other methods
Table 4 compares figures of merit of the developed method with some other SPE methods reported for the separation, preconcentration and determination of chromium species.As can be seen the LOD of the proposed method is lower or comparable and its enhancement factor is higher than the rest of methods in the Table 4.
Optimization of the SPE variablesThe pH of sample solution plays an important role on the separation/preconcentration and speciation of metal ions by SPE.It has a unique role on the metal-chelate formation, its chemical stability and the lipophilicity of the chelate that must be extracted by the IL modified MNPs sorbent.Furthermore, as in the weak acidic solution the predominant forms of Cr VI are CrO 4 2− and Cr 2 O 7 2−

Table 1 .
Temperature program of ETAAS for determination of Cr VI

Table 2 .
Tolerance limits of foreign ions for the determination of chromium(VI)

Table 3 .
Determination and speciation of chromium in real samples (n = 3) a Values in parentheses are coefficient of variation (%); b ND = not detected.

Table 4 .
Comparison of analytical characteristic of the proposed method with some SPE published method for speciation of chromium a Limit of detection (LOD); b coefficient of variation (CV); c enhancement factor (EF); d flame atomic absorption spectroscopy (FAAS); e electrothermal atomic absorption spectrometry (ETAAS).