Chromatography‐free screening and quantification of saffron adulteration with safflower by DART‐TQ

Saffron is a high value spice obtained from drying the stigmas of Crocus sativus L., mainly produced in India, Iran and other Mediterranean countries. It is classified as the most expensive spice with prices of $1,100‐11,000 per kg and therefore a prime target for fraudsters. The most common substances illegally added to saffron are powders from turmeric, safflower and paprika as well as natural and artificial dyes. Different methods have been developed for detecting saffron adulteration. However, they all have difficulties to detect adulterants like safflower added to a level lower than 20% (w/w). Here, we developed a rapid, robust, high‐throughput DART method for the reliable quantification of saffron adulteration with low safflower contents on a triple quad MS. Pure saffron was extracted by adding 5mL of EtOH/H2O at a 70/30 ratio to 50 mg of powder. After proper sample preparation, a serial addition of 10 ‐ 70% safflower to the saffron extract was performed. An EVOQ‐DART TQ+ (Bruker) was used for the analysis equipped with a JumpShot HT DART source. With statistical analyses, potential molecular markers tracing for safflowers in saffron were detected in MS full scan mode. For these selected candidates, MRM transitions were optimized. DART parameter settings like temperature, voltages and pressure were investigated as well. The repeatability of the analysis and the influence of an internal standard for the MRM method were examined. First, full scan MS spectra of saffron and safflower were acquired. Several m/z values were found to be different in intensity. E.g., a clear correlation of the peak area of both the precursor at m/z 116 and the MRM transition to m/z 70 with increasing addition of safflower was observed. An additional HR MS/MS untargeted analysis allowed us to putatively attribute the ion at m/z 116 to proline. For the quantification of these marker ions, the usage of caffeine as an internal stand significantly improved the results by compensating for the errors and variability of the measurement due to the ionization process, the sample position on the DART grid or MS related factors. The linearity of the calibration curve for the addition of 10 ‐ 70% safflower to the saffron extract showed an R2 = 0,9755. The MRM transitions monitored simultaneously along the data acquisition were m/z 116/70, 446/116 and for caffeine 195/138. For repeatability studies, two levels of 20 and 50% of safflower were spiked into the saffron extract. For 10 replicates, the CV of the recovery was 7% using caffeine as internal standard. Finally, some real cases of saffron adulteration with safflower were tested. A simple workflow was applied with a fast extraction of the sample with MeOH/H20 and dilution followed by the DART‐MRM analysis. The data which gave correct assignments of safflower down to a 5% addition with an R2 = 0,9955. The total run time for 30 samples was just 15 min.