Compound-specific carbon and hydrogen isotope analysis of volatile organic compounds using headspace solid-phase microextraction

Abstract

Natural flavouring materials are in high demand, and a premium price is paid for all-natural flavourings, making them vulnerable to fraud. At present, compound-specific isotope analysis (CSIA) is perhaps the most sophisticated tool for determining flavour authenticity. Despite promising results, the method is not widely used, and the results are limited to the most common volatile organic compounds (VOCs). This paper describes a robust protocol for on-line measurements of δ13C and δ2H using HS-SPME coupled with GC-C-IRMS and GC-HTC-IRMS for common fruit VOCs. To achieve reproducible and accurate results, a combination of a peak size/linearity correction with drift correction were used. Finally, the results were normalised by multiple point linear regression using the known and measured values of reference materials. Special care was taken to avoid irreproducible isotopic fractionation and the effects of equilibration, adsorption, desorption times and temperatures on δ13C or δ2H values were examined. Method validation was performed, and the average combined measurement uncertainty (MU) was 0.42‰. All the δ13CVPDB values were below ±3*MU, regardless of analytical conditions. In contrast, for δ2HVSMOW-SLAP values, only low temperature (30 °C) with equilibration time (15 min) and shorter adsorption time (between 10 and 20 min) can produce an isotopic difference of <10‰. Therefore, method optimisation can minimise MU, and data normalisation and method validation are essential for obtaining meaningful data for use in flavour authenticity studies.