The method used to determine tobacco-specific nitrosamines [TSNA], N’-nitrosonornicotine [NNN], 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone [NNK], N’-nitrosoanabasine [NAB] and N’-nitrosoanatabine [NAT], in mainstream smoke was modified. The methodology included trapping mainstream smoke in a buffer solution and on a Cambridge filter, extracting TSNA from the buffer solution and the filter with dichloromethane, clean-up procedures and analysis by GC/TEA. Extracts from both the buffer solution and the filter were cleaned up and analysed individually. The clean-up procedure included column chromatography on basic alumina and the removal of nicotine. N-Nitrosodibenzylamine [NDBenzA] which had suitable chromatographic properties was used as internal standard. N’-Nitrosopentylpicolylamine [NPePicA], a newly synthesized nitrosamine with physicochemical properties similar to the TSNA, could also be used as an internal standard. The recoveries for the buffer solution were as follows: NDBenzA 89 %, NPePicA 83 %, NNN 85 %, NNK 88 %, NAB/NAT 85 %. For the TSNA trapped on the filter the following recoveries were determined: NDBenzA 81 %, NPePicA 75 %, NNN 77 %, NNK 78 %, NAB/NAT 75 %. Parameters influencing the TSNA values in the mainstream smoke were investigated. Artifact formation of TSNA during mainstream smoke collection did not occur under standard conditions. TSNA values were greatly influenced by the puff volume, they increased with increasing puff volume. Reproducible and reliable results can be achieved in a relatively short time with the method reported.
