Application of strontium isotope abundance ratios measured by MC-ICP-MS for food authentication

Abstract

Naturally occurring isotopes of such elements as strontium (Sr) have proved to be good tools for detecting trends in the soil-vegetation system and the tracing of a variety of objects. Multiple-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) has been used for the precise determination of variations in the isotopic composition of Sr. The method described has been applied to the establishment of the potential and limits to determine the geographical origins of different Emmental-type cheese samples. The sample preparation consists of (i) a freeze-drying procedure to remove water, (ii) an extraction step to eliminate the fat components and to obtain the cheese casein fraction, (iii) a thermal decomposition of the latter, and (iv) a chromatographic matrix separation of the redissolved residue. The determination of the isotope abundance ratios ⁸⁸Sr/⁸⁶Sr, ⁸⁷Sr/⁸⁶Sr and ⁸⁴Sr/⁸⁶Sr resulted in precisions of 0.002–0.01%. Simultaneously, the ion currents for krypton (⁸³Kr, ⁸²Kr) and rubidium (⁸⁵Rb) were measured to correct for interferences with the Sr isotopes 84, 86 and 87. These and further (argide) spectral interferences causing bias effects to the Sr isotope abundance ratios have been investigated and an adequate computational correction procedure has been assessed. The whole set of validation data has been used for the calculation of the combined standard measurement uncertainty of the isotopic abundance ratio, resulting in a value of 0.016%. Comparison of the measured ⁸⁷Sr/⁸⁶Sr data with thermal ionisation mass spectrometric (TIMS) results, determined on the same cheese samples, agreed within the stated measurement uncertainties, thus indicating that both the validation of the sample preparation procedures and the mass spectrometric measurements cause no evident bias effect with respect to the Sr isotope abundance values. The ⁸⁷Sr/⁸⁶Sr isotope abundance ratios in cheese originating from different regions (alpine, pre-alpine, Bretagne, Finland, Canada, Australia) accorded to local geological properties. No difference was found between “casein-bound” and “whole-cheese” Sr isotope abundance ratios within the stated measurement uncertainties.