- Email: [email protected]
Method for the determination of chromium in feed matrix by HPLC
PROCESSING AND PRODUCTS Method for the determination of chromium in feed matrix by HPLC Balakrishnan Umesh,∗,1 Rajendra Moorthy Rajendran,∗ and Muthu Tamizh Manoharan† ∗
Research and Development, Kemin Industries South Asia Pvt. Ltd., India, The Trapezium, 39, Nelson Manickam Road, Chennai, Tamil Nadu, India 600029; and † Interdisciplinary School of Indian System of Medicine, SRM University, Kattankulathur 603203, Tamil Nadu, India ABSTRACT An improved method for the chromatographic separation and determination of chromium (III) and (VI) [Cr(III) and Cr(VI)] in mineral mixtures and feed samples has been developed. The method uses precolumn derivatization using ammonium pyrrolidinedithiocarbamate (APD) followed by reversedphase liquid chromatography to separate the chromium ions. Both Cr(III) and Cr(VI) species are chelated with ammonium pyrrolidinedithiocarbamate prior to separation by mixing with acetonitrile and 0.5 mmol acetate buffer (pH 4.5). Optimum chromatographic separations were obtained with a polymer-based reversed-phase column (Kinetex, 5 μ, 250 × 4.5 mm, Phenomenex, Torrance, CA) and a mobile phase containing acetonitrile and water (7:3). Both Cr(III) and Cr(VI) ion concentrations were directly determined from the corresponding areas in the chromatogram. The effect of analytical pa-
rameters, including pH, concentration of ligand, incubation temperature, and mobile phase, was optimized for both chromium complexes. The range of the procedure was found to be linear for Cr(III) and Cr(VI) concentrations between 0.125 and 4 μg/mL (r2 = 0.9926) and 0.1 and 3.0 μg/mL (r2 = 0.9983), respectively. Precision was evaluated by replicate analysis in which the percentage relative standard deviation values for chromium complex were found to be below 4.0. The recoveries obtained (85–115%) for both Cr(III) and Cr(VI) complexes indicated the accuracy of the developed method. The degradation products, as well as the excipients, were well resolved from the chromium complex peak in the chromatogram. Finally, the new method proved to be suitable for routine analysis of Cr(III) and Cr(VI) species in raw materials, mineral mixtures, and feed samples.
Key words: total chromium, HPLC validation, mineral mixtures, hexavalent chromium, chromium speciation 2015 Poultry Science 94:2805–2815 http://dx.doi.org/10.3382/ps/pev238
INTRODUCTION Some transition metals can induce adverse actions in a biological system with different toxicity expressions depending on their oxidation state. Chromium is considered to be a bio-element in the trivalent form (Anderson, 1995) or to have mutagenic properties in the hexavalent form (Katz and Salem, 1993). Chromium is an essential mineral in animal nutrition and is used in feed supplements for various applications, such as glucose homeostasis, growth performance, and antidepressant effects (Anderson and Kozlovsky, 1985; Mazzer et al., 2007; Rajalekshmi et al., 2014). The hexavalent form of chromium is highly toxic and, thus, is permissible at a level of less than 0.1 mg/L in drinking water, according to the United States Environmental Protection Agency (US EPA). Hence, it is essential to quantify both ions at trace levels in all products. C 2015 Poultry Science Association Inc. Received September 17, 2014. Accepted July 9, 2015. 1 Corresponding author: [email protected]
Chromium propionate and chromium picolinate are the two organic forms of chromium permitted by the U.S. Food and Drug Administration (FDA) for addition to swine diets up to 0.2 mg/kg, whereas in cattle feed chromium propionate is the only approved chromium source at a dosage of 0.5 mg/kg by the FDA. Hence, most chromium-based feed supplements contain low levels of the Cr(III) ion, along with stabilizers, fillers, silica, and other ingredients. There have been a large number of reports over the past few years with regard to specific analytical methods for the speciation of chromium ions individually in water (Ashraf et al., 2006). At present, atomic absorption spectroscopy (Sperling et al., 1992), inductively coupled plasma (ICP) and ion chromatography (Inoue et al., 1995) are widely used for solid matrix analysis. Wang (2010) developed a method for simultaneously quantifying chromium ions using inductively coupled plasma-mass spectrometry (ICP-MS) in urine from chromate workers. However, high-performance liquid chromatography (HPLC) is one of the most accurate, cost-effective, and widely used techniques in the analytical lab. By developing an HPLC method to quantify
2805
2806
UMESH ET AL.
Cr(III) and Cr(VI) ions simultaneously in solid matrices, the time and cost of the analysis will be reduced. Several reports have been published on using reversed-phase high-performance liquid chromatography (RP-HPLC) techniques for estimating chromium ion concentration in aqueous media. The majority of these methods have used ion exchange resins for preconcentration and co-precipitation for quantification. Kaur and Malik (2009) proposed a simultaneous determination method of chromium compounds by RPHPLC using morpholine-4-carbodithioate ligands with UV detection. The method has a broad dynamic range but is rather time consuming. Analysis of Cr(III) and Cr(VI) by co-precipitation using Fe(III) oxide and extracted with ammonium pyrrolidine thiocarbamate (Mullins, 1984). Several reports have been published on chelation extraction procedures using sodium pyrrolidinethiocarbamate ligands for simultaneous determination of chromium species in water (Tande et al., 1980; Subramanian, 1988). The RP-HPLC method was validated for the determination of chromium ions using ammonium pyrrolidine (APD) ligands in a photodiode array (PDA) detector in water samples (Hossain et al., 2005). However, none of these methods were found to be optimized for the analysis of chromium in solid matrices like mineral mixtures or feed supplements for better recoveries. This is due primarily to the interference of other ions and tedious experimental methods for extraction. Thus, a precolumn derivatization procedure using ammonium pyrrolidinethiocarbamate ligands for selective chromium estimation was attempted with an approach modified from the reported procedure (Tande et al., 1980). To optimize the HPLC condition, both complexes were individually isolated and the structures were confirmed using Fourier transform infrared spectroscopy (FTIR), electron paramagnetic resonance (EPR), and liquid chromatography–mass spectrometry (LC-MS). The complexes of Cr(III)-APD and Cr(VI)-POS have different types of polar attraction with organic solvents, but they have similar UV absorption characteristics. Therefore, Cr(VI)-POS and Cr(III)-APD complexes can be separated from their mixtures with RP-HPLC using a suitable composition of organic and aqueous mobile phases. In this paper, a comprehensive study of the chelation mechanism of chromium ions, their structural characterization, and their chromatographic separation using RP-HPLC is described. The parameters of the chelation process, including time, temperature, pH, buffer, and substrate concentration, are discussed. For quantification using HPLC, parameters such as linearity, specificity, precision, accuracy, and robustness are validated in accordance with the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) guidelines. The developed method was applied to analyze total chromium and hexavalent chromium in several solid matrices, including feed samples and mineral mixtures.
MATERIALS AND METHODS Reagents All organic solvents were of analytical grade available on the market. Acetonitrile and acetic acid were HPLC grade from Sigma-Aldrich (Bangalore, India). Deionized water was of HPLC grade quality filtered through a 0.2 micron filter from Fisher Scientific. All HPLC solvents were degassed with an ultrasonic bath prior to use. Chromium chloride hexahydrate (99.9%), potassium dichromate (99.9%), and ammonium pyrrolidinedithiocarbamate (APD, 99.8%) were obtained from Sigma-Aldrich (Bangalore, India). Sodium acetate trihydrate (Merck, India) was of laboratory grade quality and used without further purification. Mineral mixture products containing chromium (ChromflexTM C, KemTRACETM Chromium) were obtained from Kemin Industries South Asia Pvt. Ltd. (Chennai, India). The feed sample tested was a broiler finisher feed, formulated with maize, soya, wheat bran, silica, dicalcium phosphate (DCP), and minerals, including copper, manganese, zinc, cobalt, and iron.
Physical Measurements The Shimadzu Prominence HPLC unit consisted of an LC20AC solvent delivery pump and an SPD-20A UV photodiode array detector, interfaced with LCsolution Software (ver 1.25). A sample volume of 20 μL was injected into the injection valve. A Phenomenex Kinetex C18 reversed-phase column of 4.6 × 250 mm filled with C18 material with 5 μm packing was used for analysis. The column temperature was maintained at 40◦ C. A Mettler Toledo pH meter (Model S220) was used in the study for buffer preparation. The molecular weight of both chromium chelate complexes was analyzed using an Agilent 1100 series LC-MS system equipped with a dual spray electron ionization system. The stretching and bending vibrations of both complexes was analyzed using a Perkin Elmer 65 FTIR series instrument using KBr discs. Proton nuclear magnetic resonance (1 H-NMR) analysis of complexes were recorded using a 500 MHz Bruker Avance AVIII NMR Spectrometer, using d6 -DMSO as a solvent and TMS as an internal standard. Powder EPR spectra were recorded using a JEOL JES-FA200 operating at an X-band frequency (8.75–9.65 GHz) at room temperature. The magnetic moments of these complexes were recorded at room temperature using an Auto magnetic susceptibility balance (MSB) (Sherwood Scientific).
Standards/Procedure All standards and solutions were prepared using double distilled water filtered through 0.2 μ filter. APD reagent solution was prepared by dissolving 0.3 g in 100 mL water. A standard solution of 0.5 μg/mL of Cr(III) was prepared by diluting the stock solution
VALIDATION PARAMETERS
of CrCl3. 6H2 O (10.5 mg in 50.0 mL) at 40 μg/mL of Cr(III) in deionized water. Cr(VI) stock solution was prepared by dissolving potassium dichromate (25.5 mg in 50.0 mL) in distilled water to get a solution of 180.3 μg/mL of Cr(VI). A standard solution of 0.1 μg/mL Cr(VI) was prepared from Cr(VI) stock solution in distilled water. Calibration standards containing chromium species at different concentrations were prepared approximately 30 min before analysis by appropriate dilution of the Cr(III) and Cr(VI) stock solutions in glass standard flasks using distilled water and were kept at ambient temperature.
HPLC Assay of Cr Complex The metal complexation was achieved in situ for both chromium ions. The required volume of Cr stock solution was pipetted into a 10.0 mL flask. To this were added APD reagent solution (0.3%, 3.0 mL) and acetate buffer (3.0 mL) solution of pH 4.5. The flask was kept for incubation at 57◦ C for 15–20 min (Hossain et al., 2005). The solution turned turbid, which indicates chelation of the metal complex formation. The amount of precipitate formed is directly proportional to the concentration of chromium ions. The flask was allowed to cool and the precipitate was dissolved using acetonitrile and made up to volume using the same solvent in a standard 10.0 mL flask. The solution was directly injected in HPLC.
Analysis in Solid Matrices Cr analysis in chromium propionate complex. A chromium propionate-APD complex solution was prepared by adding 25 mg chromium propionate to a 100.0 mL volumetric flask with 20.0 mL of water. To this was added 1.0 mL 50% concentrated HCl solution, along with 5.0 mL water, and the mixture was heated for 10 min at 110 ± 10◦ C. The stock solution was allowed to cool and then made up to 100.0 mL using distilled water. Then 250 μL stock solution was diluted to 10.0 mL in a volumetric flask and chelation with APD was prepared as described previously. Cr analysis in mineral mixtures. A mineral mixture sample (500 mg) was weighed in a silica crucible. Three mL of 50% concentrated HCl solution was added, and the sample was digested at 100◦ C for 10 min on a hot plate. The sample solution was filtered and diluted to 50.0 mL using distilled water. Then 200 μL of solution was diluted to 10.0 mL in a volumetric flask and chelation was completed as described earlier. Cr analysis in feed samples. Feed sample (2g) was spiked with 1.0 mL of 20.0 μg/mL chromium standard solution and extracted using a 1:1 methanol:water (20.0 mL) system at 50◦ C, 150 rpm in a shaking incubator. Then the material was filtered, and 2.0 mL filtrate was taken for analysis in a 10.0 mL volumetric flask and derivatized with APD as described earlier.
2807
Method Validation The method was validated according to the ICH guidelines through the determination of range, linearity, precision, specificity, accuracy, robustness, limit of detection (LOD), and limit of quantitation (LOQ). Linearity and range. The purpose of this part of the study was to establish the linearity of the proposed method for Cr(III) and Cr(VI). Six separate series of solutions of Cr(III) (0.125–4 μg/mL) and Cr(VI) (0.10– 3.5 μg/mL) were prepared from the stock solutions and analyzed (n = 6). The calibration curve was obtained by plotting the peak area to the concentration of ions, and least squares regression analysis was performed on the obtained data. Precision. The precision of a method expresses the closeness of agreement between a series of measurements from the same homogeneous sample. A known concentration of Cr(III) and Cr(VI) ions was prepared, and both repeatability and reproducibility were evaluated. The repeatability of the method was checked by injecting 0.5 μg/mL solution of Cr(III) and Cr(VI) standards (n = 6) into HPLC. The variability of the method was studied by analyzing the solution 9 times on the same day (intraday precision) and 6 times on 3 different days (interday precision). Specificity. To check the specificity of the proposed method, a standard mixture of Cr(III) and Cr(VI) was prepared with excipients like silica, calcium carbonate, organic acids, and glycol that are commonly found in chromium propionate preparations. A comparison of excipient mixture chromatograms with the chromatograms of the standard solution was made along with the percentage recovery of both analytes. Accuracy. The accuracy of an analytical method is the closeness of test results obtained by developed method to the true value. This approach was based on the percentage relative error and mean percentage recovery of spiked Cr ions in chromium mixtures. A known concentration (0.8–1.2 μg/mL) of Cr ion was spiked with 1.0 μg/mL organic chromium chelate, and the total recovery was calculated. For Cr(III), the area corresponding to total Cr was measured in triplicate and percentage recovery was back-calculated. For Cr(VI), percentage recovery was cross verified using other analytical methods (ion chromatography) owing to lower concentrations in feed raw materials. Limit of quantitation (LOQ) and limit of detection (LOD). The LOQ and LOD parameters were determined on the basis of the signal-to-noise ratio of the analytes in the chromatogram. The repeatability of the Cr(III)-APD and Cr(VI)-POS complexes at a particular concentration should be less than 5% for LOQ and less than 10% for LOD. Robustness. The robustness of an analytical procedure refers to its capability to remain unaffected by small and deliberate variations in method parameters. The robustness of the method was studied by changing the composition of the organic phase by ±5%, pH
2808
UMESH ET AL.
by ±0.2, and using different C18 columns. Also, variation in the incubation temperature (±10%) and the concentration of ligands (±0.1%) were evaluated.
RESUTS AND DISCUSSION Characterization of Chromium Complex The reaction between chromium salts and APD and its application to chromium detection in an aqueous medium have been extensively studied (Schemes 1 and 2). However, the mechanism of complex formation and its geometry are not well understood. The objective of the present work was in part to study the complex reaction between chromium ions and APD, thereby optimizing the experimental conditions for maximum detection and reproducibility in HPLC. Electronic spectra. The electronic spectra for the Cr(III)-APD complex recorded in water gave 4 bands at 240, 254, 317, and 664 nm. The low-intensity band observed at 664 nm could be attributed to a forbidden d-d transition. The ligand-to-metal charge transition (LMCT) band at 254 and 317 nm can be attributed to a S→Cr charge transfer. The proposed structure was further confirmed by ultraviolet-visible (UV-Vis) spectral analysis. The high shift in the λmax in the UV region is a good indication for complex formation.
In the case of dichromate, chromium, which is in the hexavalent state, is reduced to the trivalent state by an APDC ligand to form a dithioperoxycarbamate complex (bis[N,N-pyrrolidine(dithioperoxycarbamateS,S )] [N,N-pyrrolidine(dithioperoxycarbamato-O,S)] chromium(III)), shown in Scheme 2 (Hossain et al., 2005). Peaks corresponding to Cr(VI) at 352 and 441 nm were shifted to longer wavelengths (497 and 636 nm), indicating the formation of a trivalent chromium complex. Also, a peak at 254 and 285 nm confirms the LMCT complex (Figure 1). The assignment of spectral bands and structural geometry for both complexes were made based on earlier reports (Hossain et al., 2005; Setiyanto et al., 2006). spectra. The IR spectrum of a ligand and metal complex is shown in Figure 2. The stretching vibration of the mercapto group (SH) of ligands is highly polarizable and tends to give a very weak signal at 2,710 cm−1 . The disappearance of this band in the spectra of both chromium complexes confirmed the ligand complex formation. The deprotonation of the SH group in chelation is confirmed by a blue shift of the υ (CS) band stretching from 835 cm−1 to 825 cm−1 . Significant change in the wavelength to a higher frequency has been observed upon complexation with a metal ion for other bands stretching in the fingerprint region. Pyrrolidine vibration bands at 1,409 cm−1 and 1,048 cm−1 of chromium complexes (Figure 2a, b) shifted to a higher
Scheme 1. Reaction of chromium chloride and ammonium pyrrolidinedithiocarbamate.
Scheme 2. Reaction of potassium dichromate and ammonium pyrrolidinedithiocarbamate.
VALIDATION PARAMETERS
Figure 1. UV-visible spectra of two chromium complexes and ligand for maximum absorbance determination.
frequency, and a change in shape was observed indicating the chelation of this group with the metal ion. There is a small shift in the bands of Cr(III)-APD and Cr(VI)POS complexes in the spectra, indicating the formation of a similar coordination complex. The disappearance of the peak at 890 cm−1 corresponds to Cr-O of dichromate (Figure 2c) and confirms the reduction of Cr(VI) to Cr(III) for the Cr-POS reaction (Hope et al., 1977). 1 H NMR spectra. The 1 H NMR spectra of the complexes show resonances that are clearly broadened and shifted by the paramagnetic chromium (III) metal center. The magnetic moments of these complexes were recorded at room temperature using an Auto-MSB. The μeff value of chromium (III) complexes ranges between 3.5 and 3.8 B.M. This further supports the idea that the complexes are paramagnetic corresponding to 3 unpaired electrons, which supports the trivalent state of chromium. Powder EPR spectra were recorded using
2809
a JEOL-JES-FA200 operating at nX-band frequency (8.75–9.65 GHz) at room temperature. EPR analysis was studied to evaluate the geometry of the metal complex shown in Figure 3b. The metal complex was synthesized and isolated in aqueous medium based on the reported procedure (Tande et al., 1980). For an octahedral d3 electron configuration of chromium (III), the ground-state electron term is 4 A2g . The g-values were calculated using the expression g = 2.0023(1–4λ/10 Dq), where λ is the spin orbit coupling constant for the metal ion. Typical experimental g-values for Cr(III) compounds having a 6-coordinated octahedral geometry range from 1.95 to 1.98 (Sulekh et al., 2001). LC-MS analysis. The mass spectra for both chromium complexes were studied in electrospray ionization mass spectrometry (ESI-MS) analysis. The molecular weight of the Cr(III)-APD complex from the mass spectra was found to be m/z 493.80 (Figure 4a), which confirmed the proposed structure (Scheme 1). Also, there was another peak at 562.05, indicating the presence of two chloride ions as a secondary valence in the coordination sphere, which was confirmed by M+2 isotopic chloride ion peaks in the spectra, whereas in the case of Cr(VI)-POS, the peroxo complex formation was confirmed from the m/z peak at 506.95 (Figure 4b). The increase in molecular weight for Cr(VI)-POS confirms the oxide bond coordination in the pyrrolidine(dithioperoxycarbamato-O,S) chromium (III) complex. The mass spectral results match the proposed structure in Figure 4, and the results are tabulated in Table 1.
Method Development and Optimization Optimization of the HPLC method is necessary for detection and reproducibility at lower levels. This method is based on precolumn derivatization using
Figure 2. FTIR spectra of a) Cr(III)-APD, b) Cr(VI)-POS, c) K2 Cr2 O7 , and d) APD.
2810
UMESH ET AL.
Figure 3. (a) 1 H-NMR, (b) EPR electronic spectra of Cr(III)-APD, and (c) Cr(VI) complexes.
APD ligands. Hence, both the HPLC method and sample preparation are critical for better reproducibility. The complexation reaction of Cr(III) and Cr(VI) with APD are highly dependent on pH. The effect of pH on complex formation is shown in Figure 5, where the difference in response for both ions at different pH conditions was observed; it was mainly due to the rate of dissociation and complex formation of ions with APD. In the case of Cr(VI), depending on the pH of the solution, it can exist as either CrO4 2− or Cr2 O7 2− . At a low pH (4–5), only the CrO4 2− ion exists, and it reacts with water to form hydrogen chromate (HCrO4− ) and then forms a complex with APD (Scheme 2). In the case of the positively charged Cr(III) ion, complexation was high at acidic pH 4.5 and attained saturation at pH 4.75 (Hossain et al., 2005). Optimization of the HPLC method was completed by varying the mobile phase and analytical column to obtain maximum separation and well-resolved symmetrical peaks. Satisfactory separation and well-resolved symmetrical peaks were obtained with the mobile phase
containing acetonitrile and water (70:30 v/v) using a Kinetex C18, 5 μ column. Other parameters, such as incubation time, oven temperature, and flow rate, were also studied; the optimized parameters are listed in Table 2. System suitability. The suitability of the HPLC instrument is very critical for Cr analysis. The suitability test parameters are tabulated in Table 3. In the aforementioned HPLC conditions, the elution of the Cr(III) and Cr(VI) complexes was obtained at a very high resolution of 13.8 min with high symmetry. The relative standard deviation (RSD) for all of the system suitability parameters was less than 5%. The retention time of Cr(VI) and Cr(III) was found to be 9.5 min and 13.0 min, respectively (Figure 6).
Validation Studies Following optimization of the HPLC conditions, the method was validated in accordance with ICH guidelines.
2811
VALIDATION PARAMETERS
Figure 4. Mass spectra of a) Cr-APD complex and b) Cr-POS complex to calculate molecular weight of both complexes.
Table 1. Results of chromium complex. Sample 1
Molecular weight
Percentage yield Percentage purity λ max (nm)
Cr-APD complex
Cr-POS complex
Theoretical: 493 Found: 493.8 82 75 254
Theoretical: 508 Found: 506.8 85 87 256
1 Mass spectral analysis with electrospray ionization mass spectrometry (ESI-MS).
Figure 5. Effect of pH on Cr-APD complex formation.
Table 2. Optimized chromatographic conditions used in this study. Parameter Chromatograph Column Mobile phase Flow rate Detection Oven temperature Injection volume
Optimized condition Shimadzu Prominence LC20A Kinetex C18, 5μ , 250×4.6 mm Acetonitrile:Water (70:30) 0.6 mL/min 254 nm, PDA detector 50◦ C 20 μ L
Linearity and range. The linearity of measurements was evaluated by analyzing different concentrations of the standard solutions of Cr(III) and Cr(VI) samples. The Beer–Lambert concentration was found to be 0.125–4 μg/mL for Cr(III) and 0.1–3.5 μg/mL for Cr(VI). A calibration curve was constructed by plotting the average peak area against concentrations, and a regression equation was computed. The r2 value of Cr(III) and Cr(VI) was 0.9926 and 0.9983, respectively, as shown in Figure 7. This method demonstrated very good linearity for both ions at trace levels. The correlation coefficient values were found to be within the
2812
UMESH ET AL.
Table 3. System suitability test parameters. Cr(III)-APD complex System suitability parameter Retention time (tR ) Area Height (mAU) Theoretical plate number (N) Tailing factor (t) ∗
Cr(VI)-POS complex
Mean
% RSD
Mean∗
% RSD
13.16 861,214 90,317 26,077
0.16 4.57 4.59 0.96
9.5 1,534,667 212,529 23,808
0.12 4.74 4.42 3.33
1.07
0.22
1.01
1.38
Mean of five replicate injections (n = 5).
guideline limits for both complexes. The results showed that an excellent correlation exists between the peak area and concentration of ions within the concentration range indicated in the graph shown in Figure 7.
Precision. An intraday precision study of the CrAPD complex of both ions was carried out by estimating the correspondence responses 9 times on the same day with 0.5 μg/mL concentration of the chromium complex. An interday precision study was conducted by analyzing the correspondence responses 5 times on the next 2 d at the same concentration. The repeatability results were found to be 3.06 and 1.82% for Cr(III) and Cr(VI) for intraday precision and 3.29 and 3.14% for interday precision, respectively. The results obtained for interday precision (% RSD) and intraday precision (% RSD) for both Cr(III) and Cr(VI) were found to be less than 5% (Table 4). Specificity. The specificity of the analysis was studied by the resolution of the two components that eluted closest to each other. Also, the peak purity test analysis using a PDA detector confirmed that the chromatographic peak for both chromium complexes could not be attributed to more than one component (Figure 6).
Figure 6. System suitability parameters of chromium complexes.
Figure 7. Linearity regression plot for a) Cr(III)-APD complex and b) Cr(VI)-POS complex.
2813
VALIDATION PARAMETERS Table 4. Precision studies of Cr (III) & Cr (VI). % RSD Cr ion
Concentration Cr complex (μ g/mL)
Cr(III) Cr(VI)
0.5 0.5
Day 1∗ Day 2† Day 3† 3.061 1.817
4.50 3.68
Average % RSD
2.31 3.92
3.29 3.14
Total number of replicates: (n = 9)∗ and (n = 6)† .
A three-dimensional chromatographic view of both complexes gives a single chromophoric signal, which confirms that there was no co-elution of another impurity peak with the analyte. Commonly used excipients (glycol, propionic acid) were spiked, and noninterference was confirmed in the chromatogram. Also, forced degradation of chromium propionate by acid and alkali hydrolysis was completed using 50% hydrochloric acid and 1N sodium hydroxide solution. There were no significant changes in the elution pattern, and no new peak was observed in the chromatogram. This confirms that this reagent is highly specific to Cr ions. The interference of other minerals, such as Zn, Cu, Cd, and Fe, was studied, and it was found that no complexation reaction with APD occurred. Accuracy. Accuracy studies were completed by spiking known the concentration of the Cr standard solution with the chromium propionate (CrP) solution. Percentage recovery was calculated based on the total chromium content. The recovery studies were carried out in triplicate over the specified concentration range, and the amount of Cr(III)-APD was estimated by measuring the peak area ratios by fitting these values to the straight line equation of the calibration curve. Three different concentrations (80–120%) of the authentic standards were added to the 1 μg/mL solution of chromium propionate. The resulting sample solutions
were injected, and chromatograms were recorded. The percentage recovery range was 94–106%, well within the ±15% specification limit. The highest percentage recovery of Cr(III)-APD was found to be 100.6%, indicating that the proposed method is highly accurate. The mean percentage recovery obtained for the total chromium is tabulated in Table 5. For the Cr-POS complex, the percentage recovery was cross verified by another analytical method using ion chromatography; the results are shown in Table 6. The measurement by ion chromatography is due to the low concentration of Cr(VI) in chromium propionate, which is well below the quantification limit. From the foregoing determination, the percentage recovery and the standard deviation of percentage recovery were calculated. The RSD value for both complexes was less than 5%, which revealed that the developed method is precise. Limit of detection/limit of quantification. The LOD and LOQ for Cr(III) were found to be 0.15 μg/mL and 0.35 μg/mL, respectively, which indicates the sensitivity of the method. Similarly, for Cr(VI), LOD and LOQ were found to be 0.07 μg/mL and 0.25 μg/mL, respectively. The LOD value clearly suggests that the developed method can be applied to quantify chromium ions at trace levels in all the samples. The results were within the specification limit. The data are tabulated in Table 7. Robustness. Robustness was analyzed by varying experimental parameters such as incubation temperature and time with respect to complex formation. No significant variation was observed at 10% deviation from the temperature (52–62◦ C) and time (15–25 min). Also, the robustness of the HPLC method was determined by making slight changes to the chromatographic conditions, such as a change in the mobile phase
Table 5. Accuracy reading of Cr (III) by recovery studies in chromium propionate. CrP concentration (μ g/mL)
Concentration of spiked Cr(III) (μ g/mL)
Total area (1μ g/mL of spiked CrP+ concentration)
Total concentration of Cr(III) found (μ g/mL)∗
Relative percentage difference
1 1 1 1 1 1 1 1 1
0.8 0.8 0.8 1.0 1.0 1.0 1.2 1.2 1.2
1,825,457 1,707,388 1,802,451 2,020,616 1,968,854 2,002,428 2,295,049 2,347,161 2,465,136
1.82 1.71 1.80 2.02 1.97 2.00 2.29 2.34 2.46
101.34 94.78 100.06 100.96 98.38 100.05 104.26 106.63 111.99
∗
Correlation coefficient: r2 = 0.9925; equation for regression line: Y = 1E + 06x + 1355.1 (n = 3).
Table 6. Comparison of results of Cr(VI) ion estimation in CrP product using HPLC and ion chromatography methods. Sample
Ion chromatography (μ g/mL)
HPLC method (μ g/mL)∗
Relative percentage difference
Cr(VI) Cr(VI)
0.29 0.26
0.26 0.28
89.65 107.9
∗
Mean replicate of three injections (n = 3).
2814
UMESH ET AL. Table 7. LOD/LOQ of Cr complex. Sample Cr(III)-APD Cr(VI)- POS ∗
LOD (μ g/mL)
LOQ (μ g/mL)
Signal/noise
% RSD∗
0.15 0.07
0.35 0.25
11.1 8.8
4.1 4.4
Mean replicate of six injections for LOQ (n = 6).
Table 8. Validation parameters. Parameter Absorption maxima Linearity range (μ g/mL) Standard regression equation Correlation Coeffiecient (r2 ) Accuracy (% recovery ± SD) Precision Intraday precision (% RSD) Interday precision (% RSD) Specificity (% RSD) LOD (μ g/mL) LOQ (μ g/mL)
Cr(III)-APD
Cr(VI)-POS
254 0.125–4 y = 4,684.8x – 14,269 0.9926 99 ± 7
254 0.1–3 y = 675,387x + 12,804 0.9983 99 ± 10
3.06 3.29 4.87 0.15 0.35
1.82 3.14 0.07 0.25
Table 9. Analysis of total Cr in feed samples/mineral mixtures. Sample Feed sample 1 Feed sample 2 KemTRACE Cra Cr blendb a b
AAS (μ g/mL)
HPLC (μ g/mL)
Relative percentage difference
1.8 1.5 8.3 2.15
1.63 1.48 7.65 2.13
90.56 98.6 92.1 99.06
kemTRACE Cr contains 5 mineral supplements, vitamins, silica. Mixture contains CrP-118 mg, 1g CaCO3 , and silica.
composition, flow rate, and column temperature by 5%. A slight shift in the retention time to 1.5 min was observed, with similar reproducibility for both ions. This method is highly robust with a slight variation in experimental conditions. No marked differences were observed in the chromatogram’s elution pattern. The complete validation results for both chromium complexes are tabulated in Table 8.
Application to Mineral Mixtures/Feed Samples The HPLC method developed in this study was used to determine the total chromium in mineral mixtures and feed samples. Mineral mixtures containing the Cr(III) ion were digested, using 50% HCl, on a hot plate for 10 min at 100◦ C. The sample was then filtered in Whatman Grade 1 filter paper (Sigma-Aldrich, Bangalore, India), and recovery was done in accordance with the general procedure. In the case of feed samples, a known concentration of Cr(VI) was spiked for recovery analysis. The feed samples were extracted using a 1:1 methanol water system at 50◦ C and 150 rpm. The sample was filtered and centrifuged, and analysis was completed in a manner similar to the general procedure of chromium analysis. The percentage recovery of chromium in feed (n = 3) and mineral mixtures was well within the specification limit, and the results are tabulated in Table 9. Also, in the mineral mixtures the
interference of fillers and other excipients (CaCO3 and silica) in the Cr blend did not interfere with the analyte concentration, and percentage recovery was calculated (Table 9). The results of the feed matrix sample and mineral mixtures analyzed show that the developed procedure is useful for determining both Cr(VI) and Cr(III) in solid matrices. The interference of other inorganic ions with APD was found to be very minimal and favors the application of this method as a quality control tool for the quantification of chromium.
CONCLUSIONS In the present study, the complexation reaction of Cr(III) and Cr(VI) with APD was briefly discussed, and the HPLC method was validated for quantification. This was the first attempt to quantify chromium ions in a solid matrix like organic chelates, mineral mixtures, and feed samples. The method was validated with acceptable performance of linearity, precision, repeatability, accuracy, and robustness in accordance with ICH guidelines. More importantly, the optimized method was successfully applied to analyze the chromium content in all the formulations.
ACKNOWLEDGMENTS The authors thank SAIF, Indian Institute of Technology, Madras, India, for help with the NMR and EPR
VALIDATION PARAMETERS
measurements and Dr. Mitch Poss and Dr. Rick Myers for their critical reading of the manuscript.
REFERENCES Anderson, R. A. 1995. Chromium, glucose tolerance, diabetes and lipid metabolism. J. Adv. Med. 8:37–49. Anderson, R. A., and A. S. Kozlovsky. 1985. Chromium intake, absorption and excretion of subjects consuming self-selected diets. Am. J. Clin. Nutr. 41:1177–1183. Ashraf, M. A., T. M. Ahmed, M. H. Zakaria, and F. Khaled. 2006. A novel kinetic determination of dissolved chromium species in natural and industrial waste water. Talanta, 70:460–467. Hope, J. M., R. Martin, L. D. Taylor, and A. H. White. 1977. Ring expansion in a metal-dithiocarbamate complex by oxygen insertion; synthesis and properties of [Cr(S2CNi(OS2CNR2))]. The X-ray structure of bis[NN-diethyl(dithiocarbamatoSS )][NN-diethyl(dithioperoxy-carbamato-OS)]-chromium(III). J.Chem.Soc. Chem. Commun. 1275:99–100. Hossain, M. A., M. Kumita, Y. Michigami, T. S. A. Islam, and S. Mori. 2005. Rapid speciation of Cr(VI) and Cr(III) by reversed-phase high-performance liquid chromatography with UV-detection. J. Chromatogr. Sci. 43:98–103. Inoue, Y., T. Sakai, and T. Kumagai. 1995. Simultaneous determination of chromium (III) and chromium (VI) by ion chromatography with inductively coupled plasma mass spectrometry. J. Chromatogr. A. 706:127–136. Katz, S. A., and H. Salem. 1993. The toxicology of chromium with respect to its chemical speciation: A review. J. Appl. Toxicol. 13:217–224. Kaur, V., and A. K. Malik. 2009. Speciation of chromium metal ions by RP-HPLC. J. Chromatogr. Sci. 47:238–242.
2815
Mazzer, P. A., L. Maurmann, and R. N. Bose. 2007. Mechanisms of DNA damage and insight into mutations by chromium (VI) in the presence of glutathione. J. Inorg. Biochem. 101:44–55. Mullins, T. L. 1984. Selective separation and determination of dissolved chromium species in natural waters by atomic absorption spectrometry. Anal. Chim. Acta. 165:97–103. Rajalekshmi, M., C. Sugumar, H. Chirakkal, and S. V. Ramarao. 2014. Influence of chromium propionate on the carcass characteristics and immune response of commercial broiler birds under normal rearing conditions. Poult. Sci. 93:574–580. Setiyanto, H., R. Muhida, T. Kishi, O. Rempillo, M. Rahman, H. K. Dipojono, W. A. Dino, S. Matsumoto, and H. Kasai. 2006. J. Vacuum Society of Japan. First-principles calculations for chemical reaction between sodium diethyldithiocarbamate and transition-metal (Cr) atom to produce Cr(DDC)(3) and Cr(DDC)(2)ODDC. Jpn. J. Appl. Phys. 45:37–41. Sperling, M., X. Yin, and B. Welz. 1992. Differential determination of chromium (VI) and total chromium in natural waters using flow injection on-line separation and preconcentration electrothermal atomic absorption spectrometry. Analyst. 117:629–635. Subramanian, K. S. 1988. Determination of chromium(III) and chromium(VI) by ammonium pyrrolidinecarbodithioate-methyl isobutyl ketone-furnace atomic absorption spectrometry. Anal. Chem. 60:11–15. Sulekh, C., P. Pipil, and S. D. sharma. 2001. Synthesis, EPR and electronic spectral studies in Cr(III) and Mn(II) complexes of some oximes. Synthesis and reactivity in Inorgic and metalorganic chemistry. 31:779–794. Tande, T., J. E. Pettersen, and T. Torgrimsen. 1980. Simultaneous determination of Cr (III) and Cr (VI) in water by reversed phase HPLC, after chelating with sodium diethyldithiocarbamate. Chromatographia. 13:607–610. Wang, H. J. 2010. Using ion-pair reversed-phase HPLC ICP-MS to simultaneously determine Cr(III) and Cr(VI) in urine of chromate workers. Talanta. 81:1856–1860.
Research and Development, Kemin Industries South Asia Pvt. Ltd., India, The Trapezium, 39, Nelson Manickam Road, Chennai, Tamil Nadu, India 600029; and † Interdisciplinary School of Indian System of Medicine, SRM University, Kattankulathur 603203, Tamil Nadu, India ABSTRACT An improved method for the chromatographic separation and determination of chromium (III) and (VI) [Cr(III) and Cr(VI)] in mineral mixtures and feed samples has been developed. The method uses precolumn derivatization using ammonium pyrrolidinedithiocarbamate (APD) followed by reversedphase liquid chromatography to separate the chromium ions. Both Cr(III) and Cr(VI) species are chelated with ammonium pyrrolidinedithiocarbamate prior to separation by mixing with acetonitrile and 0.5 mmol acetate buffer (pH 4.5). Optimum chromatographic separations were obtained with a polymer-based reversed-phase column (Kinetex, 5 μ, 250 × 4.5 mm, Phenomenex, Torrance, CA) and a mobile phase containing acetonitrile and water (7:3). Both Cr(III) and Cr(VI) ion concentrations were directly determined from the corresponding areas in the chromatogram. The effect of analytical pa-
rameters, including pH, concentration of ligand, incubation temperature, and mobile phase, was optimized for both chromium complexes. The range of the procedure was found to be linear for Cr(III) and Cr(VI) concentrations between 0.125 and 4 μg/mL (r2 = 0.9926) and 0.1 and 3.0 μg/mL (r2 = 0.9983), respectively. Precision was evaluated by replicate analysis in which the percentage relative standard deviation values for chromium complex were found to be below 4.0. The recoveries obtained (85–115%) for both Cr(III) and Cr(VI) complexes indicated the accuracy of the developed method. The degradation products, as well as the excipients, were well resolved from the chromium complex peak in the chromatogram. Finally, the new method proved to be suitable for routine analysis of Cr(III) and Cr(VI) species in raw materials, mineral mixtures, and feed samples.
Key words: total chromium, HPLC validation, mineral mixtures, hexavalent chromium, chromium speciation 2015 Poultry Science 94:2805–2815 http://dx.doi.org/10.3382/ps/pev238
INTRODUCTION Some transition metals can induce adverse actions in a biological system with different toxicity expressions depending on their oxidation state. Chromium is considered to be a bio-element in the trivalent form (Anderson, 1995) or to have mutagenic properties in the hexavalent form (Katz and Salem, 1993). Chromium is an essential mineral in animal nutrition and is used in feed supplements for various applications, such as glucose homeostasis, growth performance, and antidepressant effects (Anderson and Kozlovsky, 1985; Mazzer et al., 2007; Rajalekshmi et al., 2014). The hexavalent form of chromium is highly toxic and, thus, is permissible at a level of less than 0.1 mg/L in drinking water, according to the United States Environmental Protection Agency (US EPA). Hence, it is essential to quantify both ions at trace levels in all products. C 2015 Poultry Science Association Inc. Received September 17, 2014. Accepted July 9, 2015. 1 Corresponding author: [email protected]
Chromium propionate and chromium picolinate are the two organic forms of chromium permitted by the U.S. Food and Drug Administration (FDA) for addition to swine diets up to 0.2 mg/kg, whereas in cattle feed chromium propionate is the only approved chromium source at a dosage of 0.5 mg/kg by the FDA. Hence, most chromium-based feed supplements contain low levels of the Cr(III) ion, along with stabilizers, fillers, silica, and other ingredients. There have been a large number of reports over the past few years with regard to specific analytical methods for the speciation of chromium ions individually in water (Ashraf et al., 2006). At present, atomic absorption spectroscopy (Sperling et al., 1992), inductively coupled plasma (ICP) and ion chromatography (Inoue et al., 1995) are widely used for solid matrix analysis. Wang (2010) developed a method for simultaneously quantifying chromium ions using inductively coupled plasma-mass spectrometry (ICP-MS) in urine from chromate workers. However, high-performance liquid chromatography (HPLC) is one of the most accurate, cost-effective, and widely used techniques in the analytical lab. By developing an HPLC method to quantify
2805
2806
UMESH ET AL.
Cr(III) and Cr(VI) ions simultaneously in solid matrices, the time and cost of the analysis will be reduced. Several reports have been published on using reversed-phase high-performance liquid chromatography (RP-HPLC) techniques for estimating chromium ion concentration in aqueous media. The majority of these methods have used ion exchange resins for preconcentration and co-precipitation for quantification. Kaur and Malik (2009) proposed a simultaneous determination method of chromium compounds by RPHPLC using morpholine-4-carbodithioate ligands with UV detection. The method has a broad dynamic range but is rather time consuming. Analysis of Cr(III) and Cr(VI) by co-precipitation using Fe(III) oxide and extracted with ammonium pyrrolidine thiocarbamate (Mullins, 1984). Several reports have been published on chelation extraction procedures using sodium pyrrolidinethiocarbamate ligands for simultaneous determination of chromium species in water (Tande et al., 1980; Subramanian, 1988). The RP-HPLC method was validated for the determination of chromium ions using ammonium pyrrolidine (APD) ligands in a photodiode array (PDA) detector in water samples (Hossain et al., 2005). However, none of these methods were found to be optimized for the analysis of chromium in solid matrices like mineral mixtures or feed supplements for better recoveries. This is due primarily to the interference of other ions and tedious experimental methods for extraction. Thus, a precolumn derivatization procedure using ammonium pyrrolidinethiocarbamate ligands for selective chromium estimation was attempted with an approach modified from the reported procedure (Tande et al., 1980). To optimize the HPLC condition, both complexes were individually isolated and the structures were confirmed using Fourier transform infrared spectroscopy (FTIR), electron paramagnetic resonance (EPR), and liquid chromatography–mass spectrometry (LC-MS). The complexes of Cr(III)-APD and Cr(VI)-POS have different types of polar attraction with organic solvents, but they have similar UV absorption characteristics. Therefore, Cr(VI)-POS and Cr(III)-APD complexes can be separated from their mixtures with RP-HPLC using a suitable composition of organic and aqueous mobile phases. In this paper, a comprehensive study of the chelation mechanism of chromium ions, their structural characterization, and their chromatographic separation using RP-HPLC is described. The parameters of the chelation process, including time, temperature, pH, buffer, and substrate concentration, are discussed. For quantification using HPLC, parameters such as linearity, specificity, precision, accuracy, and robustness are validated in accordance with the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) guidelines. The developed method was applied to analyze total chromium and hexavalent chromium in several solid matrices, including feed samples and mineral mixtures.
MATERIALS AND METHODS Reagents All organic solvents were of analytical grade available on the market. Acetonitrile and acetic acid were HPLC grade from Sigma-Aldrich (Bangalore, India). Deionized water was of HPLC grade quality filtered through a 0.2 micron filter from Fisher Scientific. All HPLC solvents were degassed with an ultrasonic bath prior to use. Chromium chloride hexahydrate (99.9%), potassium dichromate (99.9%), and ammonium pyrrolidinedithiocarbamate (APD, 99.8%) were obtained from Sigma-Aldrich (Bangalore, India). Sodium acetate trihydrate (Merck, India) was of laboratory grade quality and used without further purification. Mineral mixture products containing chromium (ChromflexTM C, KemTRACETM Chromium) were obtained from Kemin Industries South Asia Pvt. Ltd. (Chennai, India). The feed sample tested was a broiler finisher feed, formulated with maize, soya, wheat bran, silica, dicalcium phosphate (DCP), and minerals, including copper, manganese, zinc, cobalt, and iron.
Physical Measurements The Shimadzu Prominence HPLC unit consisted of an LC20AC solvent delivery pump and an SPD-20A UV photodiode array detector, interfaced with LCsolution Software (ver 1.25). A sample volume of 20 μL was injected into the injection valve. A Phenomenex Kinetex C18 reversed-phase column of 4.6 × 250 mm filled with C18 material with 5 μm packing was used for analysis. The column temperature was maintained at 40◦ C. A Mettler Toledo pH meter (Model S220) was used in the study for buffer preparation. The molecular weight of both chromium chelate complexes was analyzed using an Agilent 1100 series LC-MS system equipped with a dual spray electron ionization system. The stretching and bending vibrations of both complexes was analyzed using a Perkin Elmer 65 FTIR series instrument using KBr discs. Proton nuclear magnetic resonance (1 H-NMR) analysis of complexes were recorded using a 500 MHz Bruker Avance AVIII NMR Spectrometer, using d6 -DMSO as a solvent and TMS as an internal standard. Powder EPR spectra were recorded using a JEOL JES-FA200 operating at an X-band frequency (8.75–9.65 GHz) at room temperature. The magnetic moments of these complexes were recorded at room temperature using an Auto magnetic susceptibility balance (MSB) (Sherwood Scientific).
Standards/Procedure All standards and solutions were prepared using double distilled water filtered through 0.2 μ filter. APD reagent solution was prepared by dissolving 0.3 g in 100 mL water. A standard solution of 0.5 μg/mL of Cr(III) was prepared by diluting the stock solution
VALIDATION PARAMETERS
of CrCl3. 6H2 O (10.5 mg in 50.0 mL) at 40 μg/mL of Cr(III) in deionized water. Cr(VI) stock solution was prepared by dissolving potassium dichromate (25.5 mg in 50.0 mL) in distilled water to get a solution of 180.3 μg/mL of Cr(VI). A standard solution of 0.1 μg/mL Cr(VI) was prepared from Cr(VI) stock solution in distilled water. Calibration standards containing chromium species at different concentrations were prepared approximately 30 min before analysis by appropriate dilution of the Cr(III) and Cr(VI) stock solutions in glass standard flasks using distilled water and were kept at ambient temperature.
HPLC Assay of Cr Complex The metal complexation was achieved in situ for both chromium ions. The required volume of Cr stock solution was pipetted into a 10.0 mL flask. To this were added APD reagent solution (0.3%, 3.0 mL) and acetate buffer (3.0 mL) solution of pH 4.5. The flask was kept for incubation at 57◦ C for 15–20 min (Hossain et al., 2005). The solution turned turbid, which indicates chelation of the metal complex formation. The amount of precipitate formed is directly proportional to the concentration of chromium ions. The flask was allowed to cool and the precipitate was dissolved using acetonitrile and made up to volume using the same solvent in a standard 10.0 mL flask. The solution was directly injected in HPLC.
Analysis in Solid Matrices Cr analysis in chromium propionate complex. A chromium propionate-APD complex solution was prepared by adding 25 mg chromium propionate to a 100.0 mL volumetric flask with 20.0 mL of water. To this was added 1.0 mL 50% concentrated HCl solution, along with 5.0 mL water, and the mixture was heated for 10 min at 110 ± 10◦ C. The stock solution was allowed to cool and then made up to 100.0 mL using distilled water. Then 250 μL stock solution was diluted to 10.0 mL in a volumetric flask and chelation with APD was prepared as described previously. Cr analysis in mineral mixtures. A mineral mixture sample (500 mg) was weighed in a silica crucible. Three mL of 50% concentrated HCl solution was added, and the sample was digested at 100◦ C for 10 min on a hot plate. The sample solution was filtered and diluted to 50.0 mL using distilled water. Then 200 μL of solution was diluted to 10.0 mL in a volumetric flask and chelation was completed as described earlier. Cr analysis in feed samples. Feed sample (2g) was spiked with 1.0 mL of 20.0 μg/mL chromium standard solution and extracted using a 1:1 methanol:water (20.0 mL) system at 50◦ C, 150 rpm in a shaking incubator. Then the material was filtered, and 2.0 mL filtrate was taken for analysis in a 10.0 mL volumetric flask and derivatized with APD as described earlier.
2807
Method Validation The method was validated according to the ICH guidelines through the determination of range, linearity, precision, specificity, accuracy, robustness, limit of detection (LOD), and limit of quantitation (LOQ). Linearity and range. The purpose of this part of the study was to establish the linearity of the proposed method for Cr(III) and Cr(VI). Six separate series of solutions of Cr(III) (0.125–4 μg/mL) and Cr(VI) (0.10– 3.5 μg/mL) were prepared from the stock solutions and analyzed (n = 6). The calibration curve was obtained by plotting the peak area to the concentration of ions, and least squares regression analysis was performed on the obtained data. Precision. The precision of a method expresses the closeness of agreement between a series of measurements from the same homogeneous sample. A known concentration of Cr(III) and Cr(VI) ions was prepared, and both repeatability and reproducibility were evaluated. The repeatability of the method was checked by injecting 0.5 μg/mL solution of Cr(III) and Cr(VI) standards (n = 6) into HPLC. The variability of the method was studied by analyzing the solution 9 times on the same day (intraday precision) and 6 times on 3 different days (interday precision). Specificity. To check the specificity of the proposed method, a standard mixture of Cr(III) and Cr(VI) was prepared with excipients like silica, calcium carbonate, organic acids, and glycol that are commonly found in chromium propionate preparations. A comparison of excipient mixture chromatograms with the chromatograms of the standard solution was made along with the percentage recovery of both analytes. Accuracy. The accuracy of an analytical method is the closeness of test results obtained by developed method to the true value. This approach was based on the percentage relative error and mean percentage recovery of spiked Cr ions in chromium mixtures. A known concentration (0.8–1.2 μg/mL) of Cr ion was spiked with 1.0 μg/mL organic chromium chelate, and the total recovery was calculated. For Cr(III), the area corresponding to total Cr was measured in triplicate and percentage recovery was back-calculated. For Cr(VI), percentage recovery was cross verified using other analytical methods (ion chromatography) owing to lower concentrations in feed raw materials. Limit of quantitation (LOQ) and limit of detection (LOD). The LOQ and LOD parameters were determined on the basis of the signal-to-noise ratio of the analytes in the chromatogram. The repeatability of the Cr(III)-APD and Cr(VI)-POS complexes at a particular concentration should be less than 5% for LOQ and less than 10% for LOD. Robustness. The robustness of an analytical procedure refers to its capability to remain unaffected by small and deliberate variations in method parameters. The robustness of the method was studied by changing the composition of the organic phase by ±5%, pH
2808
UMESH ET AL.
by ±0.2, and using different C18 columns. Also, variation in the incubation temperature (±10%) and the concentration of ligands (±0.1%) were evaluated.
RESUTS AND DISCUSSION Characterization of Chromium Complex The reaction between chromium salts and APD and its application to chromium detection in an aqueous medium have been extensively studied (Schemes 1 and 2). However, the mechanism of complex formation and its geometry are not well understood. The objective of the present work was in part to study the complex reaction between chromium ions and APD, thereby optimizing the experimental conditions for maximum detection and reproducibility in HPLC. Electronic spectra. The electronic spectra for the Cr(III)-APD complex recorded in water gave 4 bands at 240, 254, 317, and 664 nm. The low-intensity band observed at 664 nm could be attributed to a forbidden d-d transition. The ligand-to-metal charge transition (LMCT) band at 254 and 317 nm can be attributed to a S→Cr charge transfer. The proposed structure was further confirmed by ultraviolet-visible (UV-Vis) spectral analysis. The high shift in the λmax in the UV region is a good indication for complex formation.
In the case of dichromate, chromium, which is in the hexavalent state, is reduced to the trivalent state by an APDC ligand to form a dithioperoxycarbamate complex (bis[N,N-pyrrolidine(dithioperoxycarbamateS,S )] [N,N-pyrrolidine(dithioperoxycarbamato-O,S)] chromium(III)), shown in Scheme 2 (Hossain et al., 2005). Peaks corresponding to Cr(VI) at 352 and 441 nm were shifted to longer wavelengths (497 and 636 nm), indicating the formation of a trivalent chromium complex. Also, a peak at 254 and 285 nm confirms the LMCT complex (Figure 1). The assignment of spectral bands and structural geometry for both complexes were made based on earlier reports (Hossain et al., 2005; Setiyanto et al., 2006). spectra. The IR spectrum of a ligand and metal complex is shown in Figure 2. The stretching vibration of the mercapto group (SH) of ligands is highly polarizable and tends to give a very weak signal at 2,710 cm−1 . The disappearance of this band in the spectra of both chromium complexes confirmed the ligand complex formation. The deprotonation of the SH group in chelation is confirmed by a blue shift of the υ (CS) band stretching from 835 cm−1 to 825 cm−1 . Significant change in the wavelength to a higher frequency has been observed upon complexation with a metal ion for other bands stretching in the fingerprint region. Pyrrolidine vibration bands at 1,409 cm−1 and 1,048 cm−1 of chromium complexes (Figure 2a, b) shifted to a higher
Scheme 1. Reaction of chromium chloride and ammonium pyrrolidinedithiocarbamate.
Scheme 2. Reaction of potassium dichromate and ammonium pyrrolidinedithiocarbamate.
VALIDATION PARAMETERS
Figure 1. UV-visible spectra of two chromium complexes and ligand for maximum absorbance determination.
frequency, and a change in shape was observed indicating the chelation of this group with the metal ion. There is a small shift in the bands of Cr(III)-APD and Cr(VI)POS complexes in the spectra, indicating the formation of a similar coordination complex. The disappearance of the peak at 890 cm−1 corresponds to Cr-O of dichromate (Figure 2c) and confirms the reduction of Cr(VI) to Cr(III) for the Cr-POS reaction (Hope et al., 1977). 1 H NMR spectra. The 1 H NMR spectra of the complexes show resonances that are clearly broadened and shifted by the paramagnetic chromium (III) metal center. The magnetic moments of these complexes were recorded at room temperature using an Auto-MSB. The μeff value of chromium (III) complexes ranges between 3.5 and 3.8 B.M. This further supports the idea that the complexes are paramagnetic corresponding to 3 unpaired electrons, which supports the trivalent state of chromium. Powder EPR spectra were recorded using
2809
a JEOL-JES-FA200 operating at nX-band frequency (8.75–9.65 GHz) at room temperature. EPR analysis was studied to evaluate the geometry of the metal complex shown in Figure 3b. The metal complex was synthesized and isolated in aqueous medium based on the reported procedure (Tande et al., 1980). For an octahedral d3 electron configuration of chromium (III), the ground-state electron term is 4 A2g . The g-values were calculated using the expression g = 2.0023(1–4λ/10 Dq), where λ is the spin orbit coupling constant for the metal ion. Typical experimental g-values for Cr(III) compounds having a 6-coordinated octahedral geometry range from 1.95 to 1.98 (Sulekh et al., 2001). LC-MS analysis. The mass spectra for both chromium complexes were studied in electrospray ionization mass spectrometry (ESI-MS) analysis. The molecular weight of the Cr(III)-APD complex from the mass spectra was found to be m/z 493.80 (Figure 4a), which confirmed the proposed structure (Scheme 1). Also, there was another peak at 562.05, indicating the presence of two chloride ions as a secondary valence in the coordination sphere, which was confirmed by M+2 isotopic chloride ion peaks in the spectra, whereas in the case of Cr(VI)-POS, the peroxo complex formation was confirmed from the m/z peak at 506.95 (Figure 4b). The increase in molecular weight for Cr(VI)-POS confirms the oxide bond coordination in the pyrrolidine(dithioperoxycarbamato-O,S) chromium (III) complex. The mass spectral results match the proposed structure in Figure 4, and the results are tabulated in Table 1.
Method Development and Optimization Optimization of the HPLC method is necessary for detection and reproducibility at lower levels. This method is based on precolumn derivatization using
Figure 2. FTIR spectra of a) Cr(III)-APD, b) Cr(VI)-POS, c) K2 Cr2 O7 , and d) APD.
2810
UMESH ET AL.
Figure 3. (a) 1 H-NMR, (b) EPR electronic spectra of Cr(III)-APD, and (c) Cr(VI) complexes.
APD ligands. Hence, both the HPLC method and sample preparation are critical for better reproducibility. The complexation reaction of Cr(III) and Cr(VI) with APD are highly dependent on pH. The effect of pH on complex formation is shown in Figure 5, where the difference in response for both ions at different pH conditions was observed; it was mainly due to the rate of dissociation and complex formation of ions with APD. In the case of Cr(VI), depending on the pH of the solution, it can exist as either CrO4 2− or Cr2 O7 2− . At a low pH (4–5), only the CrO4 2− ion exists, and it reacts with water to form hydrogen chromate (HCrO4− ) and then forms a complex with APD (Scheme 2). In the case of the positively charged Cr(III) ion, complexation was high at acidic pH 4.5 and attained saturation at pH 4.75 (Hossain et al., 2005). Optimization of the HPLC method was completed by varying the mobile phase and analytical column to obtain maximum separation and well-resolved symmetrical peaks. Satisfactory separation and well-resolved symmetrical peaks were obtained with the mobile phase
containing acetonitrile and water (70:30 v/v) using a Kinetex C18, 5 μ column. Other parameters, such as incubation time, oven temperature, and flow rate, were also studied; the optimized parameters are listed in Table 2. System suitability. The suitability of the HPLC instrument is very critical for Cr analysis. The suitability test parameters are tabulated in Table 3. In the aforementioned HPLC conditions, the elution of the Cr(III) and Cr(VI) complexes was obtained at a very high resolution of 13.8 min with high symmetry. The relative standard deviation (RSD) for all of the system suitability parameters was less than 5%. The retention time of Cr(VI) and Cr(III) was found to be 9.5 min and 13.0 min, respectively (Figure 6).
Validation Studies Following optimization of the HPLC conditions, the method was validated in accordance with ICH guidelines.
2811
VALIDATION PARAMETERS
Figure 4. Mass spectra of a) Cr-APD complex and b) Cr-POS complex to calculate molecular weight of both complexes.
Table 1. Results of chromium complex. Sample 1
Molecular weight
Percentage yield Percentage purity λ max (nm)
Cr-APD complex
Cr-POS complex
Theoretical: 493 Found: 493.8 82 75 254
Theoretical: 508 Found: 506.8 85 87 256
1 Mass spectral analysis with electrospray ionization mass spectrometry (ESI-MS).
Figure 5. Effect of pH on Cr-APD complex formation.
Table 2. Optimized chromatographic conditions used in this study. Parameter Chromatograph Column Mobile phase Flow rate Detection Oven temperature Injection volume
Optimized condition Shimadzu Prominence LC20A Kinetex C18, 5μ , 250×4.6 mm Acetonitrile:Water (70:30) 0.6 mL/min 254 nm, PDA detector 50◦ C 20 μ L
Linearity and range. The linearity of measurements was evaluated by analyzing different concentrations of the standard solutions of Cr(III) and Cr(VI) samples. The Beer–Lambert concentration was found to be 0.125–4 μg/mL for Cr(III) and 0.1–3.5 μg/mL for Cr(VI). A calibration curve was constructed by plotting the average peak area against concentrations, and a regression equation was computed. The r2 value of Cr(III) and Cr(VI) was 0.9926 and 0.9983, respectively, as shown in Figure 7. This method demonstrated very good linearity for both ions at trace levels. The correlation coefficient values were found to be within the
2812
UMESH ET AL.
Table 3. System suitability test parameters. Cr(III)-APD complex System suitability parameter Retention time (tR ) Area Height (mAU) Theoretical plate number (N) Tailing factor (t) ∗
Cr(VI)-POS complex
Mean
% RSD
Mean∗
% RSD
13.16 861,214 90,317 26,077
0.16 4.57 4.59 0.96
9.5 1,534,667 212,529 23,808
0.12 4.74 4.42 3.33
1.07
0.22
1.01
1.38
Mean of five replicate injections (n = 5).
guideline limits for both complexes. The results showed that an excellent correlation exists between the peak area and concentration of ions within the concentration range indicated in the graph shown in Figure 7.
Precision. An intraday precision study of the CrAPD complex of both ions was carried out by estimating the correspondence responses 9 times on the same day with 0.5 μg/mL concentration of the chromium complex. An interday precision study was conducted by analyzing the correspondence responses 5 times on the next 2 d at the same concentration. The repeatability results were found to be 3.06 and 1.82% for Cr(III) and Cr(VI) for intraday precision and 3.29 and 3.14% for interday precision, respectively. The results obtained for interday precision (% RSD) and intraday precision (% RSD) for both Cr(III) and Cr(VI) were found to be less than 5% (Table 4). Specificity. The specificity of the analysis was studied by the resolution of the two components that eluted closest to each other. Also, the peak purity test analysis using a PDA detector confirmed that the chromatographic peak for both chromium complexes could not be attributed to more than one component (Figure 6).
Figure 6. System suitability parameters of chromium complexes.
Figure 7. Linearity regression plot for a) Cr(III)-APD complex and b) Cr(VI)-POS complex.
2813
VALIDATION PARAMETERS Table 4. Precision studies of Cr (III) & Cr (VI). % RSD Cr ion
Concentration Cr complex (μ g/mL)
Cr(III) Cr(VI)
0.5 0.5
Day 1∗ Day 2† Day 3† 3.061 1.817
4.50 3.68
Average % RSD
2.31 3.92
3.29 3.14
Total number of replicates: (n = 9)∗ and (n = 6)† .
A three-dimensional chromatographic view of both complexes gives a single chromophoric signal, which confirms that there was no co-elution of another impurity peak with the analyte. Commonly used excipients (glycol, propionic acid) were spiked, and noninterference was confirmed in the chromatogram. Also, forced degradation of chromium propionate by acid and alkali hydrolysis was completed using 50% hydrochloric acid and 1N sodium hydroxide solution. There were no significant changes in the elution pattern, and no new peak was observed in the chromatogram. This confirms that this reagent is highly specific to Cr ions. The interference of other minerals, such as Zn, Cu, Cd, and Fe, was studied, and it was found that no complexation reaction with APD occurred. Accuracy. Accuracy studies were completed by spiking known the concentration of the Cr standard solution with the chromium propionate (CrP) solution. Percentage recovery was calculated based on the total chromium content. The recovery studies were carried out in triplicate over the specified concentration range, and the amount of Cr(III)-APD was estimated by measuring the peak area ratios by fitting these values to the straight line equation of the calibration curve. Three different concentrations (80–120%) of the authentic standards were added to the 1 μg/mL solution of chromium propionate. The resulting sample solutions
were injected, and chromatograms were recorded. The percentage recovery range was 94–106%, well within the ±15% specification limit. The highest percentage recovery of Cr(III)-APD was found to be 100.6%, indicating that the proposed method is highly accurate. The mean percentage recovery obtained for the total chromium is tabulated in Table 5. For the Cr-POS complex, the percentage recovery was cross verified by another analytical method using ion chromatography; the results are shown in Table 6. The measurement by ion chromatography is due to the low concentration of Cr(VI) in chromium propionate, which is well below the quantification limit. From the foregoing determination, the percentage recovery and the standard deviation of percentage recovery were calculated. The RSD value for both complexes was less than 5%, which revealed that the developed method is precise. Limit of detection/limit of quantification. The LOD and LOQ for Cr(III) were found to be 0.15 μg/mL and 0.35 μg/mL, respectively, which indicates the sensitivity of the method. Similarly, for Cr(VI), LOD and LOQ were found to be 0.07 μg/mL and 0.25 μg/mL, respectively. The LOD value clearly suggests that the developed method can be applied to quantify chromium ions at trace levels in all the samples. The results were within the specification limit. The data are tabulated in Table 7. Robustness. Robustness was analyzed by varying experimental parameters such as incubation temperature and time with respect to complex formation. No significant variation was observed at 10% deviation from the temperature (52–62◦ C) and time (15–25 min). Also, the robustness of the HPLC method was determined by making slight changes to the chromatographic conditions, such as a change in the mobile phase
Table 5. Accuracy reading of Cr (III) by recovery studies in chromium propionate. CrP concentration (μ g/mL)
Concentration of spiked Cr(III) (μ g/mL)
Total area (1μ g/mL of spiked CrP+ concentration)
Total concentration of Cr(III) found (μ g/mL)∗
Relative percentage difference
1 1 1 1 1 1 1 1 1
0.8 0.8 0.8 1.0 1.0 1.0 1.2 1.2 1.2
1,825,457 1,707,388 1,802,451 2,020,616 1,968,854 2,002,428 2,295,049 2,347,161 2,465,136
1.82 1.71 1.80 2.02 1.97 2.00 2.29 2.34 2.46
101.34 94.78 100.06 100.96 98.38 100.05 104.26 106.63 111.99
∗
Correlation coefficient: r2 = 0.9925; equation for regression line: Y = 1E + 06x + 1355.1 (n = 3).
Table 6. Comparison of results of Cr(VI) ion estimation in CrP product using HPLC and ion chromatography methods. Sample
Ion chromatography (μ g/mL)
HPLC method (μ g/mL)∗
Relative percentage difference
Cr(VI) Cr(VI)
0.29 0.26
0.26 0.28
89.65 107.9
∗
Mean replicate of three injections (n = 3).
2814
UMESH ET AL. Table 7. LOD/LOQ of Cr complex. Sample Cr(III)-APD Cr(VI)- POS ∗
LOD (μ g/mL)
LOQ (μ g/mL)
Signal/noise
% RSD∗
0.15 0.07
0.35 0.25
11.1 8.8
4.1 4.4
Mean replicate of six injections for LOQ (n = 6).
Table 8. Validation parameters. Parameter Absorption maxima Linearity range (μ g/mL) Standard regression equation Correlation Coeffiecient (r2 ) Accuracy (% recovery ± SD) Precision Intraday precision (% RSD) Interday precision (% RSD) Specificity (% RSD) LOD (μ g/mL) LOQ (μ g/mL)
Cr(III)-APD
Cr(VI)-POS
254 0.125–4 y = 4,684.8x – 14,269 0.9926 99 ± 7
254 0.1–3 y = 675,387x + 12,804 0.9983 99 ± 10
3.06 3.29 4.87 0.15 0.35
1.82 3.14 0.07 0.25
Table 9. Analysis of total Cr in feed samples/mineral mixtures. Sample Feed sample 1 Feed sample 2 KemTRACE Cra Cr blendb a b
AAS (μ g/mL)
HPLC (μ g/mL)
Relative percentage difference
1.8 1.5 8.3 2.15
1.63 1.48 7.65 2.13
90.56 98.6 92.1 99.06
kemTRACE Cr contains 5 mineral supplements, vitamins, silica. Mixture contains CrP-118 mg, 1g CaCO3 , and silica.
composition, flow rate, and column temperature by 5%. A slight shift in the retention time to 1.5 min was observed, with similar reproducibility for both ions. This method is highly robust with a slight variation in experimental conditions. No marked differences were observed in the chromatogram’s elution pattern. The complete validation results for both chromium complexes are tabulated in Table 8.
Application to Mineral Mixtures/Feed Samples The HPLC method developed in this study was used to determine the total chromium in mineral mixtures and feed samples. Mineral mixtures containing the Cr(III) ion were digested, using 50% HCl, on a hot plate for 10 min at 100◦ C. The sample was then filtered in Whatman Grade 1 filter paper (Sigma-Aldrich, Bangalore, India), and recovery was done in accordance with the general procedure. In the case of feed samples, a known concentration of Cr(VI) was spiked for recovery analysis. The feed samples were extracted using a 1:1 methanol water system at 50◦ C and 150 rpm. The sample was filtered and centrifuged, and analysis was completed in a manner similar to the general procedure of chromium analysis. The percentage recovery of chromium in feed (n = 3) and mineral mixtures was well within the specification limit, and the results are tabulated in Table 9. Also, in the mineral mixtures the
interference of fillers and other excipients (CaCO3 and silica) in the Cr blend did not interfere with the analyte concentration, and percentage recovery was calculated (Table 9). The results of the feed matrix sample and mineral mixtures analyzed show that the developed procedure is useful for determining both Cr(VI) and Cr(III) in solid matrices. The interference of other inorganic ions with APD was found to be very minimal and favors the application of this method as a quality control tool for the quantification of chromium.
CONCLUSIONS In the present study, the complexation reaction of Cr(III) and Cr(VI) with APD was briefly discussed, and the HPLC method was validated for quantification. This was the first attempt to quantify chromium ions in a solid matrix like organic chelates, mineral mixtures, and feed samples. The method was validated with acceptable performance of linearity, precision, repeatability, accuracy, and robustness in accordance with ICH guidelines. More importantly, the optimized method was successfully applied to analyze the chromium content in all the formulations.
ACKNOWLEDGMENTS The authors thank SAIF, Indian Institute of Technology, Madras, India, for help with the NMR and EPR
VALIDATION PARAMETERS
measurements and Dr. Mitch Poss and Dr. Rick Myers for their critical reading of the manuscript.
REFERENCES Anderson, R. A. 1995. Chromium, glucose tolerance, diabetes and lipid metabolism. J. Adv. Med. 8:37–49. Anderson, R. A., and A. S. Kozlovsky. 1985. Chromium intake, absorption and excretion of subjects consuming self-selected diets. Am. J. Clin. Nutr. 41:1177–1183. Ashraf, M. A., T. M. Ahmed, M. H. Zakaria, and F. Khaled. 2006. A novel kinetic determination of dissolved chromium species in natural and industrial waste water. Talanta, 70:460–467. Hope, J. M., R. Martin, L. D. Taylor, and A. H. White. 1977. Ring expansion in a metal-dithiocarbamate complex by oxygen insertion; synthesis and properties of [Cr(S2CNi(OS2CNR2))]. The X-ray structure of bis[NN-diethyl(dithiocarbamatoSS )][NN-diethyl(dithioperoxy-carbamato-OS)]-chromium(III). J.Chem.Soc. Chem. Commun. 1275:99–100. Hossain, M. A., M. Kumita, Y. Michigami, T. S. A. Islam, and S. Mori. 2005. Rapid speciation of Cr(VI) and Cr(III) by reversed-phase high-performance liquid chromatography with UV-detection. J. Chromatogr. Sci. 43:98–103. Inoue, Y., T. Sakai, and T. Kumagai. 1995. Simultaneous determination of chromium (III) and chromium (VI) by ion chromatography with inductively coupled plasma mass spectrometry. J. Chromatogr. A. 706:127–136. Katz, S. A., and H. Salem. 1993. The toxicology of chromium with respect to its chemical speciation: A review. J. Appl. Toxicol. 13:217–224. Kaur, V., and A. K. Malik. 2009. Speciation of chromium metal ions by RP-HPLC. J. Chromatogr. Sci. 47:238–242.
2815
Mazzer, P. A., L. Maurmann, and R. N. Bose. 2007. Mechanisms of DNA damage and insight into mutations by chromium (VI) in the presence of glutathione. J. Inorg. Biochem. 101:44–55. Mullins, T. L. 1984. Selective separation and determination of dissolved chromium species in natural waters by atomic absorption spectrometry. Anal. Chim. Acta. 165:97–103. Rajalekshmi, M., C. Sugumar, H. Chirakkal, and S. V. Ramarao. 2014. Influence of chromium propionate on the carcass characteristics and immune response of commercial broiler birds under normal rearing conditions. Poult. Sci. 93:574–580. Setiyanto, H., R. Muhida, T. Kishi, O. Rempillo, M. Rahman, H. K. Dipojono, W. A. Dino, S. Matsumoto, and H. Kasai. 2006. J. Vacuum Society of Japan. First-principles calculations for chemical reaction between sodium diethyldithiocarbamate and transition-metal (Cr) atom to produce Cr(DDC)(3) and Cr(DDC)(2)ODDC. Jpn. J. Appl. Phys. 45:37–41. Sperling, M., X. Yin, and B. Welz. 1992. Differential determination of chromium (VI) and total chromium in natural waters using flow injection on-line separation and preconcentration electrothermal atomic absorption spectrometry. Analyst. 117:629–635. Subramanian, K. S. 1988. Determination of chromium(III) and chromium(VI) by ammonium pyrrolidinecarbodithioate-methyl isobutyl ketone-furnace atomic absorption spectrometry. Anal. Chem. 60:11–15. Sulekh, C., P. Pipil, and S. D. sharma. 2001. Synthesis, EPR and electronic spectral studies in Cr(III) and Mn(II) complexes of some oximes. Synthesis and reactivity in Inorgic and metalorganic chemistry. 31:779–794. Tande, T., J. E. Pettersen, and T. Torgrimsen. 1980. Simultaneous determination of Cr (III) and Cr (VI) in water by reversed phase HPLC, after chelating with sodium diethyldithiocarbamate. Chromatographia. 13:607–610. Wang, H. J. 2010. Using ion-pair reversed-phase HPLC ICP-MS to simultaneously determine Cr(III) and Cr(VI) in urine of chromate workers. Talanta. 81:1856–1860.