The pyrolysis of non-volatile tobacco ingredients using a system that simulates cigarette combustion conditions

This is the second part of a systematic study in which tobacco ingredients are pyrolysed using experimental conditions designed to simulate the average combustion conditions inside a burning cigarette. In the first part, the pyrolysis system was developed and single-substance, mostly semi-volatile tobacco ingredients were pyrolysed. It was predicted that on a cigarette, the majority of these semi-volatile ingredients would transfer to smoke with little pyrolysis. In this part of the study, a further 159 non-volatile and complex ingredients, as well as ingredient mixtures, have been pyrolysed and the pyrolysis products determined using a gas chromatography-mass spectrometric system coupled to the pyrolyser. These non-volatile tobacco ingredients generally decomposed completely in the pyrolysis system, often yielding many products in relatively small amounts. The study has concentrated on the biologically active substances produced by pyrolysis, in particular the Hoffmann analytes. These analytes are believed by regulatory authorities in Canada and U.S.A. to be relevant to smoking-related diseases. They are based on lists published by Hoffmann and co-workers of the American Health Foundation in New York. For the pyrolysis of many of the non-volatile ingredients, no Hoffmann analytes were detected amongst the products. When they were occasionally formed, they included phenols, benzene, toluene, styrene and furfural (furfural is biologically active but it does not appear on any of the Hoffmann or regulatory authority lists). Those ingredients that did yield such products generally produced them in relatively small quantities although furfural was produced in relatively large quantities by pyrolysis of some ingredients, especially sugars. Those ingredients that produced biologically active constituents during their pyrolysis have been further assessed. This was done by adding them to cigarettes, machine-smoking the cigarette and comparing their smoke yields to those from a control (no ingredient) cigarette. From this comparison, it was found that in general the ingredients added to cigarettes do not increase the smoke components relative to the control cigarette. The pyrolysis technique of the present study tends to over-predict the amount of decomposition that the non-volatile ingredients undergo relative to their behaviour in a burning cigarette. Several examples are discussed, in particular ingredients that produce furfural during pyrolysis. This general pyrolysis technique is thus a first step in the total toxicological assessment of tobacco ingredients and is a useful screening tool for indicating which ingredients may yield biologically active products during decomposition of the ingredients. There are, however, some products such as formaldehyde and the carbon oxides that are not detected by the pyrolyser-gas chromatography-mass spectrometric technique employed here. The generation and detection of these products during the pyrolysis of selected tobacco ingredients is the subject of a parallel paper. (C) 2005 Elsevier B.V. All rights reserved.