Exploration of CF3+ as a new reagent ion for Chemical-Ionization-Reaction Mass-Spectrometry
2019-01-29T15:04:18Z (GMT) by
The detection of small alkanes in a chemical ionization reaction mass spectrometer (CIR-MS) using conventional reagents presents an analytical challenge. The proton affinity of reagents such as H3O+ is too high, whereas the alternatives, such as O2+, give rise to excessive fragmentation. In this research, the use of CF3+, generated from CF4, in CIR-MS is shown to be suitable to detect light n-alkanes in the range C2-C6 with almost no fragmentation. CF3+ ionization proceeds via hydride abstract ion and this is the critical difference in comparison with the proton transfer mechanism facilitated by H3O+. Different reaction mechanisms are observed to be at work in the reaction of CF3+ with a functionally varied range of volatile organic compounds (VOCs). For example, with the larger alkanes, fluoride extraction occurs and, with aromatic and nitrile VOCs, CF3+ acts as a Lewis acid and a strong electrophile. Cases of charge transfer were observed as well as a metathesis reaction with aldehydes and ketones, where the oxygen in the C=O bond is replaced with C-F+. The weakness of the bonds in longer chain VOCs were found to exhibit the same tendency to fragment as experienced with reactions with every other type of reagent, including H3O+. Application of CF3+ as a reagent, in exhaled breath and urine headspace analysis of smokers and non-smokers was examined. The results showed that CF3+ was more sensitive than H3O+ in the detection of acetonitrile and toluene in exhaled breath. In both cases of acetonitrile and toluene, an adduct is formed with CF3+, which makes these molecules easily identifiable from interfering isobaric compounds. CF3+ reactions were not affected by the large amounts of urea and ammonia found in urine samples. When using H3O+ as a reagent, VOCs with lower proton affinities such as small n-alkanes cannot be detected. Hence CF3+ has the potential to be a useful as a complementary reagent to H3O+ to detect a wider range of VOCs in real-life samples. The data of breath and urine headspace were subject to multivariate analysis, mainly Principle Component Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLS-DA) which provided excellent means for distinguishing between smokers and non-smokers in breath and urine analysis.