Analysis of Toxic Trace Elements in Coffee Bean Products by HR ICP-OES

Transcription

1 Analysis of Toxic Trace Elements in Coffee Bean Products by HR ICP-OES Abstract Samples of coffee powder and caffeine extracts from a coffee bean processor were submitted to the analysis of As, Cd, Cr, Cu, Hg, Mn, Ni and Pb traces by HR ICP-OES on equipped with Standard Kit. Samples were analyzed after microwave digestion of 0.5 g in a mixture of 6 ml HCl and 2 ml HNO 3 with a total sample volume of 50 ml. Aqueous calibration standards were prepared from 1000 ppm single elements standards (SIGMA ALDRICH). For particular elements of interest multiple elemental lines have been used to show method robustness. Reasons for previous Cadmium overestimation are discussed in detail. Briefly, elevated Iron and Arsenic contents present in some coffee bean products gave rise to spectral interference on Cadmium emission lines used in the past (conventional ICP-OES). In contrast, the unique spectral resolution of the High Resolution Array ICP-OES provides FWHM values of typically less than 3 pm for respective Cadmium lines, so that spectral interferences by Iron and Arsenic can be ruled out. Challenge Analysis of toxic trace elements in coffee and caffeine extracts Matrix Coffee powder digests, caffeine extracts Purpose Clarification of high Cadmium contents 1/6

2 Materials and Methods Calibration Table 1: Concentration of calibration standards Element Unit Cal.1 Cal.2 Cal.3 Cal.4 Cal.5 As Cd Cr Cu Hg Mn Ni Pb Method Parameters Table 2: Plasma configurations and set-up of the sample introduction system Parameter Specification Power 1200 W Plasma gas flow 12 L/min Auxillary gas flow 0.5 L/min Nebulizer gas flow 0.5 L/min Nebulizer Concentric nebulizer (borosilicate, 1.0 ml/min) Spray chamber Cyclonic Spray Chamber, 50 ml, borosilicate Outer tube/inner tube Quartz / Quartz Injector Quartz, 2 mm Pump tubing PVC Sample pump rate 1.0 ml/min Rinse/ Read delay 30 s 2/6

3 Evaluation Parameters Table 3: Overview of method-specific evaluation parameters Element Line [nm] Plasma -view Integration- Mode Read time [s] Evaluation No. of pixel Baseline fit Polynomial degree Correction As axial peak 3 3 ABC 1 auto - As axial peak 3 3 ABC auto - Cd axial peak 3 3 ABC auto - Cd axial peak 3 3 ABC auto - Cr axial peak 3 3 ABC auto - Cr axial peak 3 3 ABC auto - Cu axial peak 3 3 ABC auto - Cu axial peak 3 3 ABC auto - Hg axial peak 3 3 ABC auto - Mn axial peak 3 3 ABC auto - Mn axial peak 3 3 ABC auto - Ni axial peak 3 3 ABC auto - Ni axial peak 3 3 ABC auto - Pb axial peak 3 3 ABC auto - 1 automatic baseline correction (ABC) 3/6

4 Results and Discussion Table 4: Overview of results for coffee powder sample Element Line Expected concentration Measured concentration RSD Spike recovery DL 1 [nm] As / As / Cd / Cd / Cr / Cr / Cu / Cu / Hg unknown <DL 0.81 Mn / Mn / Ni / Ni / Pb / detection limit (DL) from blank method (with respect to initial sample weight in digestion) 4/6

5 Table 5: Overview of results for caffeine extract Element Line Expected concentration Measured concentration RSD Spike recovery DL 1 [nm] As / As / Cd / Cd / Cr / Cr / Cu / Cu / Hg unknown <DL 0.81 Mn / Mn / Ni / Ni / Pb / detection limit (DL) from blank method (with respect to initial sample weight in digestion) Clarification of previous Cadmium overestimation For both samples it was noted that the Cadmium levels established by HR ICP-OES were significantly lower than the values expected by the customer. That is, on both Cd and Cd nm gave Cadmium levels of about 50, while the customer had reported significantly higher Cadmium contents in the range from 0.25 and 1.0 mg/l in the past. Close analysis revealed that Cadmium lines at and nm were routinely used at the customer site that are affected by spectral interferences from Iron and Arsenic on conventional ICP-OES, respectively. Hence, insufficient separation of the emission lines from the analyte (Cd) and these common interferences (Fe, As) result in false contribution of the latter to the Cadmium content in the sample. 5/6

6 Figure 1: Comparison of the Cadmium/Arsenic line pair recorded on ICP-OES instruments with conventional (red) and high spectral resolution (green). This becomes clear when comparing the spectrum of the Cadmium/Arsenic line pair about nm asgiven in Figure 1. On conventional ICP-OES instruments the emission lines from Cadmium and Arsenic traces are indistinguishable (red spectrum), while the High Resolution Array ICP-OES PlasmaQuant PQ 9000 Elite allows for both emission lines to be well-separated (green spectrum). Only in the latter case will the analysis of the Cadmium line be unaffected by the Arsenic line in close vicinity, so that correct results are assured. Reference: ICP_OES_36_14_en.docx This document is true and correct at the time of publication; the information within is subject to change. Other documents may supersede this document, including technical modifications and corrections. Content may be used without written permission but with citation of source Publisher: 6/6