Chromium Speciation Using an Aminated Amberlite XAD-4 Resin Column Combined with Microsample Injection-Flame Atomic Absorption Spectrometry

Amberlite XAD-4 resin (AXAD-4) was chemically modified to an aminated Amberlite XAD-4 (AAXAD-4) resin and characterized by infrared spectroscopy. AAXAD-4 resin was used as an efficient solid phase for the preconcentration and speciation of Cr(III) and Cr(VI) ions by column technique. The concentration of chromium species was determined by microsample injection system-flame atomic absorption spectrometer (MIS-FAAS). Selective retention of Cr(III) ions was achieved at pH 8.0 and eluted using 1.0 mL of 3.0 mol L–1 HCl and 1.0 mL of 2.0 mol L–1 NaOH, successively, at the flow rate of 5.0 mL min–1. The maximal sorption capacity of AAXAD-4 resin for Cr(III) ions was found to be 67.0 mg g–1. The limit of detection (LOD) and limit of quantitation (LOQ) for Cr(III) ions were found to be 0.041 and 0.131 μg L–1, respectively, with preconcentration factor (PF) of 375 and relative standard deviation (RSD) of 3.75% (n = 11). The method was validated using certified reference materials (CRMs) and successfully applied to the real samples, spiked with Cr(III) and Cr(VI) ions. Keyword: Aminated Amberlite XAD-4 resin; column; solid-phase; chromium speciation; MIS-FAAS


Introduction
Speciation of chromium is still one of most important long-standing analytical challenges due to its impact on environmental chemistry, ecotoxicology, clinical toxicology and food industry. Among several redox states, chromium exists mostly in the trivalent Cr(III) and hexavalent Cr(VI) redox states with contrasting chemical, biological and toxicological properties. While water insoluble Cr(III) is an essential ion for mammals, water soluble Cr(VI) is a human carcinogen, mutagen and toxin due to its high oxidation potential and relatively small size. Compounds of Cr(VI) are 10 to 100 times more toxic than those of Cr(III). 1,2 Cr(III) and Cr(VI) also cause dermatologic allergy during contact. Thus, US EPA and WHO recommend the threshold value for total chromium as 100 μg L -1 for drinking water and 50 μg L -1 Cr(VI) as tolerance level, respectively. 3,4 The toxicity of metals strongly depends on their oxidation states rather than their total concentrations. 5 Therefore, metallic species have become a prime task for analytical chemists for years. 6 Various techniques, such as flame atomic absorption spectrometry (FAAS), graphite furnace atomic absorption spectrometry (GF-AAS), inductively coupled plasma mass spectrometry (ICP-MS), inductively coupled plasma atomic emission spectrometry (ICP-AES), inductively coupled plasma optical emission spectrometry (ICP-OES), thermospray flame furnace atomic absorption spectrometry (TS-FF-AAS) and electrothermal atomic absorption spectrometry (ET-AAS), have been routenly used for the determination of total chromium. 7 Unfortunately these techniques cannot differentiate Cr(III) from Cr(VI) ions. For the speciation and preconcentration of chromium species, several methods based on solid-phase extraction (SPE), [8][9][10][11] coprecipitation, [12][13][14] cloud point extraction 15 and liquid phase microextractions [16][17][18] have been developed. Among these, SPE has advantages such as easy operation, smallest consumption of toxic solvents, recycling of solid phases and high selectivity. 19,20 For speciation of chromium species, activated carbon, 21 silica gel, 22 sawdust, 23 chelating resins [11][12][13][14][15][16][17][18][23][24][25] and Amberlite XAD resin series have been used as solid phase. [25][26][27] Amine group was incorporated on the polymeric matrix of Amberlite XAD-4 resin. This modified resin was used as effective solid phase for SPE speciation and preconcentration of chromium species.

1. Apparatus
A Perkin-Elmer flame atomic absorption spectrometer (AAnalyst 200) equipped with a chromium hollow cathode lamp, an air-acetylene flame atomizer and handmade microinjection system was used for chromium determination. The instrumental parameters were established as recommended by manufacturer: wavelength, 357.9 nm; lamp current, 30.0 mA; slit width, 0.7 nm; acetylene flow, 2.0 L min -1 and air flow, 17.0 L min -1 . As reported in the previous study, a 100 μL volume (for all sample and standard solutions) was injected manually into a micropipette tip of the microinjection system connected to the nebulizer of FAAS. 13 The pH of solution was carefully measured using a digital pH meter (Hanna 211, Germany). ATR-IR spectrometer (UATR Spectrum Two from PerkinElmer) was used for recording ATR spectra. The reverse osmosis system (Human Corp., Seoul, Korea) was used to obtain ultrapure (UP) quality water (resistivity, 18.2 MΩ cm -1 ).

Reagents and Solutions
Analytical grade chemicals and UP water were used throughout the study. Stock solutions of Cr(III) and Cr(VI) were prepared using high-purity Cr(NO 3 ) 3 . 9H 2 O (Sigma-Aldrich, St. Louis, MO, USA) and K 2 Cr 2 O 7 (Merck, Darmstadt, Germany), respectively. The working and reference solutions were prepared daily by diluting the stock solutions. Amberlite XAD-4 resin was purchased from Alfa Aesar (Germany). The pH was adjusted using CH 3 COOH/CH 3 COONa buffer to pH 3-6, a solution of equal volume of 1.0 mol L -1 HCl and 1.0 mol L -1 NaOH solutions for pH 7 and NH 4 NO 3 /NH 3 buffer for pH 7.5-10.

3. Sampling
The bottled drinking and mineral water samples were purchased from a local market in Denizli, Turkey. The waste water samples were collected from outlet of the wastewater treatment plant in Denizli, Turkey. The fountain water was taken from Incilipınar, Denizli, Turkey. The waste water samples were immediately transported to the laboratory and filtered with 0.45 µm cellulose nitrate membrane (Sartorius, GmbH, Germany) under vacuum to remove suspended materials and then analysed by the proposed procedure within 24 h.

4. Chemical Modification of Amberlite XAD-4 resin
Amberlite XAD-4 resin (polystyrene divinyl benzene) was chemically modified by the reported procedure. 26,27 5.0 g of Amberlite XAD-4 resin was put into 250 mL round bottom flask and a nitrating mixture of 10 mL of concentrated HNO 3 and 25 mL of concentrated H 2 SO 4 was added. The system was stirred for 1 h at 60 o C. Reaction mixture was poured into an ice-cold water and filtered. The nitro derivative was washed repeatedly with cold water until acid was rinsed out and air-dried. The nitro group was reduced to amino derivative by refluxing with 40 g of SnCl 2 and 60 mL of 2.0 mol L -1 HCl in 100 mL of ethanol. The amino product (AAXAD-4) was treated thoroughly with 2.0 mol L -1 sodium hydroxide to decompose the tin-amine complex, followed by 1.0 mol L -1 HCl in order to remove the excess stannous chloride. Finally, the product was washed with excess water and dried at 75 o C in drying oven for 24 h. The final resin product was confirmed by infrared spectroscopy.

5. Preparation of SPE Column
A purchased empty Chromabond column SPE cartridge tube (3 mL) from Macherey-Nagel, Düren, Germany, was packed with 185 mg of aminated resin (ground). Glass wool was used to pack both ends of the column to avoid the loss of the resin during experiments. The flow rate of the sample solution was controlled with Chromabond vacuum manifold. The SPE column was decontaminated by washing with acetone, 1.0 mol L -1 HCl, 1.0 mol L -1 NaOH and water, respectively. For adjusting pH of the resin to 8, NH 4 NO 3 /NH 3 buffer was passed through the column.

6. Speciation and Preconcentration Procedure
The model solutions in the range of 10-750 mL including 5-10 µg Cr(III) or Cr(VI) were adjusted to pH 8 and passed through the column. Cr(III) ions were retained on the resin and Cr(VI) ions were passed as effluent. Cr(III) ions were eluted by sequential use of 1.0 mL of 3.0 mol L -1 HCl and 1.0 mL of 2.0 mol L -1 NaOH at the flow rate of 5.0 mL min -1 . The Cr (III) ions in the eluent were determined by MIS-FAAS. The recovery of Cr(III) ions was quantitatively achieved. The total concentration of chromium was determined by the same procedure after the reduction of Cr(VI) to Cr(III) ions using reported reducing mixture of 0.5 mL of ethanol and 0.5 mL of concentrated H 2 SO 4 . 29

1. Characterization
The modification of Amberlite XAD-4 resins was characterized by infrared spectroscopy. In supporting information, ATR-IR spectra of unmodified Amberlite XAD-4 resin ( Figure S1), nitro derivative ( Figure S2) and amino derivative ( Figure S3) are given. By comparing spectra ( Figure S1 and S2), the additional strong peaks in Figure S1 spectrum at 1525 and 1347 cm -1 , respectively, correspond to the asymmetric and symmetric stretching vibrations of N=O bond in nitro derivative. 28 By comparing spectra in Figure S2 and Figure S3, appearance of two characteristic peaks at 3359 and 3217 cm -1 in Figure S3 spectrum corresponds to N-H bond in amino derivative (primary amine). The spectral information revealed that Amberlite XAD-4 resin was successfully converted to amino derivative.

2. Effect of pH
The pH is an important parameter that strongly influences the retention of metal species on the surface of the resin. Thus, the effect of pH between 2 and 9 on the adsorption of Cr(III) and Cr(VI) ions was studied separately. For optimization, 50 mL of model solutions at pH from 2 to 9 was passed through the column individually. The adsorbed Cr(III) and Cr(VI) ions were eluted by sequential use of 2.5 mL of 3.0 mol L -1 of HCl and 2.5 mL of 2.0 mol L -1 of NaOH and determined by MIS-FAAS. At pH 7 through 9, the recoveries of Cr(III) and Cr(VI) ions were ≥95% and ≤10%, respectively, as shown in Figure 1. Therefore, pH 8 was selected as the best point for the separation of Cr(III) and Cr(VI) ions. Very low uptake of Cr(VI) ions at pH 7 through 10 can be explained as amino group of AAXAD-4 resin became negatively charged in alkaline medium and possessed electrostatic repulsion with CrO 4 2ions that caused a decrease in the uptake of Cr(VI) ions. At low pH values, Cr(III) exists as its kinetically non-reactive aqua-complex Cr(H 2 O) 3 3+ that leads to its low uptake due to possible electrostatic repulsion between Cr(H 2 O) 3 3+ and protonated amino group of AAXAD-4 resin. As pH was increasing, the coordinated water molecules were replaced by the more reactive hydroxide ions, transforming the former complex (Cr(H 2 O) 3 3+ ) to a more active form (Cr(H 2 O) 2 (OH) 2+ or Cr(H 2 O)(OH) 2 + ), which leads to comparatively better interaction with amino group (-NH 2 ) of AAXAD-4 resin. 31

Effect of Eluents
The effects of type, volume and concentration of eluents were tested for the quantitative desorption of Cr(III) ions from the column. Figure 1 clearly indicates the percentage decrease in recoveries below pH 3 for the uptakes of Cr(III) ions by AAXAD-4 resin. Thus, 5.0 mL of HCl with concentration range from 1.0 through 7.0 mol L -1 was tested to elute the Cr(III) ions. The recovery of Cr(III) was not achieved quantitavely up to 7.0 mol L -1 HCl as shown in Table 1. At pH>8.5 ( Figure 1) the uptake of Cr(III) ions decreased due to the conversion of Cr(OH) 3 to highly soluble tetrahydroxo complex (Cr(OH) 4 -). Thus, 5.0 mL of NaOH with concentration range from 1.0 through 4.0 mol L -1 was tested to elute the Cr(III) ions. The quantitative recovery of Cr(III) ions was not achieved until up to 4.0 mol L -1 NaOH (Table 1). Based on these results, a consecutive use of 2.5 mL of 3.0 mol L -1 HCl and 2.5 mL of 2.0 mol L -1 NaOH solutions was tested for desorption of Cr(III) ions and resulted in quantitative recov-  NaOH (to obtain high preconcentration factor) and resulted in quantitative recovery of Cr(III) ions (Table 1). Thus, a consecutive use of 1.0 mL of 3.0 mol L -1 HCl and 1.0 mL of 2.0 mol L -1 NaOH solutions was selected as the best eluetion solvent for the desorption of Cr(III) ions in further experiments.

4. Effect of Sample Volume
Another strategy to concentrate analyte at very low concentration is to increase the volume of sample. Therefore, the effect of sample volume on the retention of Cr(III) was studied. The recovery of Cr(III) ion was achieved quantitatively (≥95%) up to the sample volume of ≤750 mL as shown in Figure 2. Thus, the PF was calculated to be 375 as the ratio of maximal sample volume (750 mL) to minimal eluent volume (2.0 mL). Considering time factor, the volume of real samples for analysis was fixed to 100 mL.
in the range of 0.5-6.0 mL min -1 . The results (Figure 3) demonstrated that the quantitative retention and percentage recovery of Cr(III) ions were achieved at the flow rate of 5.0 mL min -1 of sample solution and eluent as well.

6. Adsorption Capacity
The adsorption capacity of the resin is a significant parameter that determines the minimal quantitity of adsorbent required for quantitative uptake of analyte from a sample solution. Based on a previous report in reference 31

5. Effect of Flow Rate of Eluent and Sample Solution
In order to decrease the preconcentration time, the flow rates of sample and eluent solutions were optimised

7. Sorption Competition of Coexisting Ions with Cr(III) Ions
Environmental water samples contain many heavy metal ions and some common alkali and alkaline earth metals as coexisting ions. For this reason, the effect of present coexisting ions on the preconcentration of Cr(III) needs to be evaluated at optimal conditions. For this purpose, 20 mL of model solution containing 1.0 µg L -1 of Cr(III) ions was spiked with possible interfering ions and subjected to the column according to the proposed method. The Cr(III) ions were quantitatively recovered in the presence of coexisting ions at tolerance limits, taken as a relative error ≤ ±5%. It can be seen from Table 2

8. Cr(III) Determination in Presence of Cr(VI) and Determination of Total Chromium Amount
The applicability of the proposed method was tested in presence of Cr(VI) ions for the determination of Cr(III) ions. For testing, the synthetic aqueous solutions including various mixtures of Cr(III) and Cr(VI) ions at different concentration levels were passed through the column at optimal conditions. Cr(III) ions were quantitatively separated and retained while Cr(VI) ions were almost completely passed through the column. This was observed by analyzing the effluent. The recoveries of Cr(III) ions were achieved quantitatively as shown in Table 3. In further study, the usability of the method for the determination of total chromium amount was also tested. Total chromium was determined after the reduction of Cr(VI) ions to Cr(III) ion by adding a mixture of 0.5 mL of concentrated H 2 SO 4 and 0.5 mL of ethanol to 50 mL of sam-ple solution containing Cr(VI) and Cr(III) ions at different concentration levels (Table 3). 29 The recovery of total chromium was also achieved quantitatively as shown in Table 3.

9. Analytical Performance of the Proposed Method
The accuracy and validation of proposed method was confirmed by analysing different CRMs such as industrial wastewater (BCR-715), drinking water (TMDW-500) and lyophilised water (BCR-544) for the determination of Cr(III) ions and total chromium. It was checked by Student's t-test whether the difference between the certified value and the found value was significant. The results shown in Table 4 indicated that there is not a significant difference between certified and found values.
After preconcentration of Cr(III) ions, the linear equation was A = 5.5259X + 0.0008 and r 2 = 0.9995 for 600 mL with concentration range of 2-12 µg L -1 of Cr(III) ions. Before preconcentration, the linear equation was A = 0.0191X + 0.0021 and r 2 = 0.998 within the concentration range of 0.2-5.0 µg mL -1 of Cr(III) ions. Theoretical PF was calculated to be 289 as the ratio of slope of linear equation after preconcentration to the slope of linear equation before preconcentration close to the experimental PF of 300, indicating the retention and eluation of the analyte was quantitative with recovery of 96%. The sensitivity was found to be 5.53 µg L -1 from the slope of the calibration curve. 32 The reproducibility of the overall precocentration method in terms of RSD was calculated to be 3.75% (n = 11) at the concentration of 0.5 µg L -1 Cr(III) ions. LOD (blank + 3σ) and LOQ (blank + 10σ, where σ is RSD of blank analysis, n = 20) are defined by IUPAC and were calculated accordingly. 33,34 The LOD and LOQ of Cr(III) ions were found to be 0.041 and 0.131µg L -1 , respectively. AAXAD-4 resin was successfully reused more than 250 times without significant loss in its performance.

10. Application of the Developed Method
The proposed method was applied successfully on different real water samples for the determination of Cr(III) ions and total chromium. The samples were analysed before and after spiking with Cr(III) ions and Cr(VI) ions. The recoveries of Cr(III) ions from the samples were achieved quantitatively as shown in Table  5. The total chromium levels of Incilipınar drinking fountain water and outlet water of waste water plant (Denizli, Turkey) samples do not pose a risk for public health.

Comparison
Analytical performance of the proposed method was compared with recently reported methods. In comparsion, LOD and PF of reported method are better than those of reported methods shown in Table 6.

Conclusion
In this work, a modified AAXAD-4 resin column was evaluated for the speciation of Cr(III) and Cr(VI) ions, providing for selective preconcentration of Cr(III) at high pH. Besides its good selectivity between Cr(III) and Cr(VI) ions, it also has some characteristics such as good stability under working conditions, fast sorption and desorption kinetics, large adsorption capacity and good tolerance to coexisting ions. The used SPE system could recover more than 95% of Cr(III) from aqueous solution at pH=8. The feasibility of speciation at µg L -1 levels make it an efficient sorbent for Cr(III). By com-bining AAXAD-4 minicolumn SPE with MIS-FAAS, the developed method was successfully applied for chromium speciation in various water samples with low LOD, high PF, good accuracy and repeatability. Because of its simplicity, low cost and safety, it could be adopted for routine use for the speciation of Cr(III) and Cr(VI) ions.

Conflict of Interest
Authors declare that they do not have any conflict of interest with anyone.

Acknowledgement
The authors would like to acknowledge the scientifical research projects unit of Pamukkale University which is the fund to this study (No. 2014 FEF 011).