This recommended practice enables the quantification of volatile compounds in flavourings to be made by gas chromatography with flame-ionization detection, without having authentic compounds available, and also in many cases it can avoid time-consuming calibration procedures. The relative-response factors (RRF) can be predicted from the molecular formula of the compound, and this approach can be applied to compounds containing the atoms C, H, O, N, S, F, Cl, Br, I, and Si, providing that the molecular formula and number of benzene rings in the analytes are known. The purity of chemicallydefined flavouring substances or chromatographic standards can also be estimated using these predicted RRF, and this procedure can also be used to quantify (poly)hydroxylated compounds, after their derivatization into trimethylsilyl ethers or esters.

The fragrant terpenes limonene and linalool can form skin sensitizing hydroperoxides upon prolonged exposure to air. Recently, high frequencies of positive patch tests to oxidized linalool and limonene were reported from multiple dermatological centres. However, there is a lack of data indicating potential sources of consumer exposure to sensitizing doses of terpene hydroperoxides which explains this frequent contact allergy. Within the IDEA project (International Dialogue for the Evaluation of Allergens; http://ideaproject.info/), a taskforce was formed to drive analytical method development and evaluation. In an inter‐laboratory study in five laboratories, a method based on hydroperoxide reduction combined with GC–MS was tested for reproducibility. Blinded samples of commercial fine fragrances were spiked with four different hydroperoxides. In samples spiked with 100–200 μg/ml, an average recovery of 86–105% with a relative standard deviation between laboratories of 7.4–22% was found. In samples spiked with 20–50 μg/ml, the recovery was 85–91%. The reduction approach offers a transferable and reproducible method to indirectly detect low levels of hydroperoxides, at least in fine fragrances. Ideally, one would prefer to directly detect the parent hydroperoxide. Therefore the same samples were further tested with three LC‐based methods directly detecting the parent hydroperoxide. LC coupled to chemiluminescence, LC‐Q‐TOF‐MS or LC‐orbitrap‐MS were used. Results indicate that with specific gradients a separation of the four analytes and quantification in the fragrance matrix can be achieved. Results of this method evaluation study present a toolbox of methods to detect terpene hydroperoxides to further investigate consumer exposure.

et al.. Exposure source for skin sensitizing hydroperoxides of limonene and linalool remains elusive: an analytical market surveillance. Food and Chemical Toxicology, Elsevier, 2019, 127, pp. AbstractReports about positive patch test reactions to oxidized linalool and limonene remain frequent. These terpenes are fragrance ingredients widely present in consumer products. The main sensitizing ingredients in the oxidation mixtures of these terpenes are hydroperoxides (HP). Currently, it is not clear whether fragranced consumer products are a relevant exposure source for HP. Analytical methodologies had been developed and validated in blind-coded ring-trials in multiple laboratories allowing quantification of the HP in different consumer products. The analytical approach had been successfully transferred to an independent third party laboratory and was now used in the analytical investigation of consumer products. In total, 104 products were analysed with a method based on hydroperoxide reduction followed by GC-MS. Samples included aged and new samples from the same brand, products which were suspected by patch test positive patients to elicit their symptoms and some products containing high levels of essential oils. Only four samples contained > 50 µg/g of at least one of four analysed HP by the reduction method. Confirmatory analysis by LC-MS methods directly testing for presence of the hydroperoxide indicated that levels are even below those observed by the conservative reduction method. The samples retrieved from patch-test positive samples were below detection limit for all four target analytes by GC-MS, and LC analysis with three methods confirmed this negative result. This independent market surveillance indicates that concentrations of HP in investigated consumer products and patient products are orders of magnitude below reported sensitizing or elicitating doses. No evidence for hydroperoxide accumulation in aged products or products used by patients could be found. The nature and source of the inducing agent responsible for the frequent positive patch test reactions to oxidized terpenes remains elusive. The analytical work with GC-MS at SOLVIAS was funded by IDEA. The analysis by LC-MS was funded by the three participating laboratories (Firmenich, IFF and Givaudan). The study management was funded by IDEA. Manuscript writing was funded by Givaudan Schweiz AG. The synthesis of reference standards and sample collection was funded by the IDEA project. We thankfully acknowledge all donors of samples, in particular the Spanish dermatological network collecting samples from patients (GEIDAC, Grupo Español de Investigación en Dermatitis Alérgica de Contacto). Other members of the IDEA HP taskforce are gratefully acknowledged for fruitful discussions during various workshops. IDEA management team: Hans J. Bender; Industry scientists: Alain Chaintreau, Hugues Brévard and Neil Owen; Scientists from Academia:

No abstract

The adoption of the 7th amendment of the European Cosmetic Directive 76/768/EEC requires any cosmetic product containing any of 26 raw materials identified by the Scientific Committee on Cosmetic Products and Non-Food Products intended for Consumers as likely to cause a contact allergy when present above certain trigger levels to be declared on the package label. Of these 26, 24 are volatile and can be analyzed by GC. This paper describes a method for the quantitative analysis of these volatile raw materials in perfume ingredients as well as complex perfume compositions. The method uses sequential dual-column GC-MS analysis. The full-scan data acquired minimize the false-positive and false-negative identifications that can be observed with alternate methods based on data acquired in the SIM mode. For each sample, allergen levels are determined on both columns sequentially, leading to two numerical results for each allergen. Quantification limits for each allergen in a perfume mixture based on the analysis of a standard are <4 mg/kg. This is well below the level that would trigger label declaration on the consumer good. Calibration curves for all allergens are linear (r > 0.999) and stable for multiple days. Studies on perfumes spiked with multiple allergens at 30, 50, and 70 mg/kg show recoveries close to nominal values.

The performances of the GC-MS determination of suspected allergens in fragrance concentrates have been investigated. The limit of quantification was experimentally determined (10 mg/L), and the variability was investigated for three different data treatment strategies: (1) two columns and three quantification ions; (2) two columns and one quantification ion; and (3) one column and three quantification ions. The first strategy best minimizes the risk of determination bias due to coelutions. This risk was evaluated by calculating the probability of coeluting a suspected allergen with perfume constituents exhibiting ions in common. For hydroxycitronellal, when using a two-column strategy, this may statistically occur more than once every 36 analyses for one ion or once every 144 analyses for three ions in common.

Simultaneous distillation-extraction (SDE) combines, in a single step, hydro-distillation of volatiles from a sample with continuous solvent extraction of the solutes dissolved in the aqueous distillate. The solvent extract containing the isolated flavour volatiles can be subsequently analysed by gas chromatography. The technique is applicable to compounds possessing a wide range of volatility, but it is suitable neither for compounds of low volatility (e.g. vanillin), nor for highly volatile materials (e.g. methanethiol). The present guidelines concern the isolation of flavouring substances by simultaneous distillation-extraction (SDE), based on the experience of the authors in laboratories of the flavour industry.

The fragrant terpenes limonene and linalool can form skin‐sensitizing hydroperoxides (HPs) upon prolonged exposure to air. Sources of exposure of consumers to sensitizing doses of HPs have not been identified, and it is not clear whether fragranced products are a relevant source. Previously this question was addressed via analytical studies on fine fragrances; however, linalool and limonene are widely used in different consumer products, especially in other leave‐on toiletries. Hence, analytical methods also need to be able to detect potential HPs in more complex consumer product matrices. Here we applied different simple extraction methods and a toolbox of analytical methods to creams and lotions. Blinded samples of a commercial skin cream and a body lotion were spiked with four different HPs at different doses. Five laboratories analysed the samples with a method based on HP reduction in the sample, followed by Extrelut® NT extraction and GC‐MS to quantify the formed alcohols. This method found an average recovery of spiked levels of 80–105%, with a relative standard deviation between laboratories of 11–25% in samples spiked with 100–200 μg g−1. Quantification was also possible in samples spiked with 20–50 μg mL−1, with a relative standard deviation between laboratories of 11–38%. Thus, this method can indirectly detect low levels of HPs in complex bases. In parallel, the same samples were analysed with three LC‐based methods directly detecting the parent HPs: LC coupled with chemiluminescence, LC‐Q‐TOF‐MS, and LC‐orbitrap‐MS. On average, the different analytes were detected with a recovery of 80–143%. No HPs were detected in the non‐spiked products, despite the fact that they do contain linalool and limonene. Results of these studies indicate that consumer exposure can now be studied routinely in different product types as the required methods are ready for roll‐out.