Volumen 21 (2004): Heft 2 (July 2004)

In the mid-1980s, each major US cigarette manufacturer prepared a list of those ingredients added at that time to its cigarette products. The lists were combined into one and submitted to the US Office of Smoking and Health in 1986. It comprised 599 entities. On the basis of extensive literature survey and examination of much unpublished data from the Tobacco Industry members on the chemistry and toxicology of the ingredients, a panel of eminent toxicologists assessed the safety of each listed ingredient with regard to its pyrolysate components and its possible effect when added to cigarette tobacco on the chemical and biological properties of the cigarette mainstream smoke. Subsequently, Doull et al. listed the 599 ingredients and summarized the conclusions of the panel on their effect on the chemical and biological properties of cigarette smoke.

In addition to the panel and Doull et al., other investigators have noted that many of the compounds used as ingredients in cigarette tobacco blends are identical with or similar to identified components of tobacco and/or tobacco smoke. The validity of this statement is obvious when the compounds in the Doull et al. list are cataloged as in Table 1. Those tobacco ingredients that are not individual compounds but are naturally-occurring oils, resins, etc. or extracts of naturally-occurring materials not only contain many of the compounds listed by Doull et al. as tobacco additives but also contain many of the same compounds present in tobacco.

Detailed examination of the literature on the chemical and biological properties of the recently used tobacco ingredients listed by Doull et al. plus a massive amount of chemical and biological data generated during the past several decades indicates that not only does none of the Doull et al. listed ingredients contribute any significant adverse chemical properties to cigarette mainstream smoke (MSS) but also none affects adversely the biological properties of the MSS.

The chemical factors examined included: a) The effect on MSS composition of the ingredients added to cigarette tobacco at the usual use level or several times that. In two major series of studies, one by Carmines et al. and one by Baker et al., the effect of the added ingredient on the concentration in mainstream smoke of specific components defined as toxicants was determined. The Carmines et al. study involved analysis of the smoke components suggested by the US Consumer Product Safety Commission and of concern to the International Agency for Research on Cancer (IARC). The Baker et al. study involved analysis of the so-called ‘Hoffmann analytes’ in cigarette smoke. b) The nature of the pyrolysis products generated during the smoking process or during pyrolysis of an individual ingredient under conditions approximating those in the cigarette pyrolysis zone. In many instances when the added ingredient is a compound, a significant percentage of it is transferred unchanged to the MSS and sidestream smoke (SSS). The small percentage not transferred intact to the smoke is seldom converted to an MSS component possessing significant toxic properties.

The extensive biological studies that showed no significant adverse effect of the MSS from ingredient-containing cigarettes included: a) The specific tumorigenicity to laboratory animal skin of the mainstream cigarette smoke condensate (CSC) from ingredient-containing cigarettes vs. the mainstream CSC from ingredient-free cigarettes. b) Exposure of laboratory animals via inhalation to the MSS from ingredient-containing cigarettes vs. the MSS from ingredient-free cigarettes. c) Determination in a variety of tests of the in vitro genotoxicity of the mainstream particulate phase and/or vapor phase.

In addition, the results of non-tobacco-related studies are available in which many individual compounds on the Doull et al. list were assayed for mutagenicity in the Ames test with several strains of Salmonella typhimurium. An excellent example is the 1984 study by Ishidate et al. who examined the mutagenicity of many compounds included as additives in Japanese foods. Over 40 of the compounds exhibiting non-mutagenicity also occur on the Doull et al. tobacco ingredient list.

Assessment of the total chemical and biological data cited herein provides a noteworthy contradiction to the much repeated assertions - with no data supporting them - that the ingredients added to cigarette tobacco result in significant adverse changes in the chemical and biological properties of the cigarette MSS.

Cigarette-burning and the smoke-formation processes and smoke composition are important topics for understanding cigarette performance. This paper proposes the molecular formulas representing the active components of bright, burley, and Oriental tobaccos and a basic chemistry model of the cigarette burning processes. Previous knowledge of the cigarette burning processes and smoke formation helped to establish parameters in deriving the basic chemistry equations. The proposed chemistry provides a brief view of the mechanisms of the cigarette burning during puffing and interpuff smoldering, and can be used to interpret and predict the smoke composition for cigarettes made from bright, burley, and Oriental tobaccos. Based on the proposed chemistry, the effect of ventilation on smoke component deliveries is discussed and the reaction heat of the puffing process is estimated.

Experiments have been conducted to examine the effect of different levels of blend glycerol (0 to 11.4 wt% on cigarette mainstream smoke (MSS) yields. The glycerol transfer is generally found to be proportional to the blend glycerol level with the maximum glycerol contribution to the nicotine-free-dry-particulate-matter (NFDPM or ‘tar’) per cigarette at 36% for a 11.4% blend glycerol. For cigarettes with different designs the glycerol in NFDPM may also depend on the glycerol loading per unit rod length. The tobacco rod filtration did not change significantly within the glycerol range investigated and hence plays a relatively minor role. Significant glycerol condensation ahead of the burning coal after a puff was measured. This condensation may have implications on glycerol levels in the sidestream smoke during inter-puff smouldering.

At the request of the UK Department of Health, samples of 25 commercial UK cigarette brands were provided to LGC Ltd a for smoke analysis. The brands reflected a high market share (58% in July 2001) and included a wide range of blend and product styles manufactured and imported into the UK.= 0.76), suggesting a minor role of other design features on constituents yield variability. This was confirmed by the application of multiple regression analysis to the data. A subset of five brands, retested at another laboratory, gave between-laboratory differences in mean constituent yields of as much as 2.5-fold. Consideration of these results, other likely sources of analytical variation in this study and a review of other studies, clearly indicates that any tolerance values to be associated with individual smoke constituent measurements will be greater than those for NFDPM, and in some cases, much greater. Consistent with the reported results from other large studies it is concluded that, under ISO smoking conditions, smoke constituent yields are largely predictable, if NFDPM and CO yields are known, for a standard cigarette. Given these observations and the likely limitations of analytical determination, the need for routine measurement of smoke constituent yields, other than NFDPM, nicotine or CO, on standard cigarettes, is questionable.

The Technical Committee ISO/TC 126 ‘Tobacco and tobacco products’ of the International Organization for Standardization set up Working Group 7 to develop and test a method for the determination of ‘tar’ and nicotine yield when fine-cut smoking articles (FCSAs) are smoked by routine-analytical smoking machines. FCSAs are individually made by consumers using tobaccos manufactured for enclosure in paper wrappers (i.e. leaflets and tubes - with or without filters). This paper records the history of method development, defines terms and explains the differences between measurements made on FCSAs and commercially factory made cigarettes. Individual consumer making practices are represented by the specification of four hand-made article types using two wrappers of different physical properties each with either 400 mg or 750 mg tobacco.

ISO standards require estimates of the repeatability (r) and reproducibility (R) of analytical methods, so the working group has conducted a collaborative study. The values ofr andR were determined using the definitions given in ISO 5725-2 (18). Twenty laboratories took part and tested three different commercial tobacco blends according to the standard method. The detailed protocol for the study is given in Annex A. Annex B is an illustrated manual used as a training aid for laboratory personnel with little experience of making FCSAs. After the smoking was completed, the laboratory data were examined for outlying points by the methods of Grubbs and Cochran. It was found necessary to exclude the data from one laboratory prior to statistical analysis otherwise the incidence of outliers was low (1.41%). Values ofr andR were calculated from the results of one-way analyses of variance (within and between laboratories) for smoke nicotine and nicotine-free dry particulate matter (NFDPM). The data for the 19 laboratories are given in Annex C. The quality of the data is high as is shown by the comparisons made between ther andR values for the articles and those for the monitor test pieces which were used to confirm that smoking conditions run-to-run were satisfactory. The range of NFDPM and smoke nicotine yields resulting from the four designs of articles is large for each of the three blends. The measured NFDPM yields of the four designs are ranked in the same order by all of the laboratories and all pairs are statistically different for every laboratory. For smoke nicotine the ranking order is the same for all laboratories. All pairs are statistically different in all but three laboratories for which only the 400 mg articles were not statistically different. Ther andR data are given in tables and in graphical representation. The graphs show thatr andR increase with the mean values of NFDPM and smoke nicotine yields. Such a trend has also been found in studies using factory-made cigarettes. A comparison of results shows thatr values for articles are approximately 40% greater andR values about 100% greater. The variability of individual hand-made articles in this study is likely to be much greater than the variability of machine-made matched cigarette samples. The method detailed in ISO 15592-3, therefore, is satisfactory for general adoption even in laboratories where there has been little previous experience.