Accuracy improvement of determination of seven minor tobacco alkaloids in mainstream cigarette smoke using analyte protectants by gas chromatography-mass spectrometry

Abstract

Tobacco alkaloids are important precursors of carcinogenic tobacco-specific nitrosamines. Therefore, accurate quantification of tobacco alkaloids is highly important. This study investigates the compensation effects of novel analyte protectants (APs) for matrix effects (MEs) to determine seven minor tobacco alkaloids (nornicotine, myosmine, anabasine, anatabine, nicotyrine, 2,3′-bipyridine, and cotinine) in mainstream cigarette smoke with high accuracy and robustness. By comparing the heights and shapes of the chromatographic peaks before and after the addition of APs to standard solutions prepared in dichloromethane and cigarette smoke solutions, the compensation effects of 12 APs and their combinations on the MEs of seven minor tobacco alkaloids were evaluated, and the best combination of 2-pyridylethylamine (2 mg/mL)+1,2-decanediol (1 mg/mL) was identified. This AP combination could effectively improve the shapes and increase the heights (by 7–371%) of chromatographic peaks for standard solutions prepared in dichloromethane and cigarette smoke solutions. Before the addition of this AP combination, the slope ratios of the calibration curves for two types of standard solutions of the seven target chemicals were 71.4–159.8%, while they were 87.4–105.6% after the addition, indicating that this AP combination reduced the matrix difference between pure solvent and cigarette smoke solution. After adding the AP combination, the standard curves of solutions prepared in dichloromethane showed good linearity (r2 ≥ 0.999), the spiked recoveries were between 80.9% and 119.6%, and the inter- and intraday precisions were between 1.5–9.5% and 3.1–8.5%, respectively. Three commercial cigarette samples and one mixed standard solution were also tested under four different instrument working conditions to verify the long-term accuracy and ruggedness of the approach in routine real-world analysis of the method. The results showed that the RSD values were higher (3.5–25.4%) without the AP combination than that (<6.7%) with the AP combination. Because of its high accuracy, precision, and robustness, this method has good practical prospects.