Volume 9 (1977): Issue 3 (October 1977)

An unsteady-state technique has been used to measure convective heat transfer coefficients in the tobacco rods of American and British cigarettes. The coefficients measured in dry cigarettes near room temperature at flow rates from 7.6 to 28 cm³/s increase linearly from 0.0003 to 0.0012 cal/(cm² s °C). Because of channeling of the gas flow in the relatively porous (70-80 %) tobacco rods, these heat transfer coefficients are an order of magnitude lower than those for large scale packed beds, in which the gas flow is more evenly distributed.

The effects of the cigarette periphery on smoke formation have been experimentally studied. A series of cigarettes was prepared with carbon-14 labelled materials placed in different radial positions within the rods. These cigarettes were smoked and the distribution of the radioactivities into total smoke was measured. These data lead to the following conclusions: [1 ] The periphery plays a large contributing role (though not the sole role) in the formation of a) the mainstream particulate phase, b) sidestream particulate phase (derived from non-distillable/non-sublimable materials), and c) mainstream gas phase (derived from non-distillable/non-sublimable materials) ; [2] The central portion of the cigarette rod plays a large contributing role in a) sidestream gas phase and b) sidestream particulate phase (derived from distillable/sublimable compounds); and [ 3 ] The mainstream gas phase yields from the distillable/sublimable compounds are due in large part to the pyrolysis/combustion products formed during the static burn period. These products are carried into mainstream smoke by the incoming air at the beginning of the puff and by diffusion due to the pressure gradient existing in the cigarette during static burn. These experiments have demonstrated the significant effects that the central portion of the cigarette rod has on sidestream smoke formation and, conversely, the effects of the periphery on mainstream smoke.

In some cigarette designs, diffusion of carbon monoxide out of a cigarette as the gases are drawn from their formation region towards the mouth end of the cigarette is more important than air dilution in controlling carbon monoxide mainstream delivery. The diffusion is a three-stage process: radial diffusion through the tobacco bed, diffusion through the paper, and diffusion away from the outer surface of the paper. Measurements have been made of the diffusion coefficient of carbon monoxide through paper, and the diffusion coefficient from the cigarette. The former has a value of 4.4 × 10^-3 cm² s^-1 for paper of permeability 70 cm³ min^-1 (10 cm²)^-1 (10 cm water)^-1, whereas the latter has the much lower values of 1.5 to 1.9 × 10^-3 cm² s^-1 (dependent on external air movement) for an unlit cigarette wrapped in the same paper. In a lit cigarette, the combined diffusion coefficient is increased by about 50 %, due to thermal effects. The diffusion of carbon monoxide away from the outer surface of the paper is a significant component of the overall diffusion coefficient determined for cigarettes. This diffusion rate is affected by the external air movement round the cigarette. Thus the draught conditions under which cigarettes are smoked could have an important influence on the carbon monoxide delivery determined. For inherently porous cigarette papers, the diffusion coefficient of carbon monoxide through the paper is approximately proportional to the square root of permeability. For a perforated paper, the diffusion coefficient is dependent on the permeability of the base (unperforated) paper, and virtually independent of the degree of perforation. These observations are explicable using a simple model of the paper structure.

The effects of diffusion and dilution on the concentration of gaseous components have been studied by a model consisting of unlit cigarettes differing in paper porosity and length, and a standard gas mixture. Results are as follows: The difference in the concentration of gaseous components between the front and butt end of the cigarette during the puff increases with cigarette length and paper porosity, and also depends on the diffusion coefficient of respective components through the cigarette paper. These changes could be explained mainly by the diffusion loss through the paper in the case of the cigarette with ordinary paper, but by both dilution with air through the paper and diffusion loss in the case of the cigarette with perforated paper. Subsequently, the levels of gaseous components such as carbon monoxide in the exit stream become lower in the Iatter cigarette. In addition, the following equation, which can fully account for these changes, has been derived:

where C1i and C0i are the concentrations of the component i at the front and butt end, respectively, L the cigarette length, r the cigarette radius, s the thickness of paper, U0 and U1 the apparent linear gas velocities at the butt and front end, respectively, and Di the diffusion coefficient of component i through the paper.

In order to clarify the thermal decomposition mechanism of tobacco cell-wall polysaccharides, some of the primary pyrolysis products were investigated using a Curie-point pyrolysis technique. The major primary products of tobacco cellulose prepared from midribs of Nicotiana tabacum, heated at 460°C for 5 seconds in helium or in vacuum, were such 1,6-anhydrosugars as levoglucosan, 1,6-anhydrob-D-glucofuranose, 1,4,3,6-dianhydro-a-D-glucopyranose and levoglucosenone, and such furan compounds as 5-hydroxymethyl-2-furaldehyde and 2-furaldehyde. The major primary pyrolysis products of tobacco xylan prepared from stalks of Nicotiana tabacum were 3-hydroxy-2-penteno-1,5-lactone and 2-furaldehyde. Some discussions of the thermal reaction mechanism were presented, based on these results.

A new method for determining particle-size distribution of cigarette smoke particles was developed by simultaneous measurement of scattered light at three angles for a fixed wavelength. A theoretical chart useful for this purpose, which was made of the relative intensities of scattered light at the angles 45° and 135° to that at the angle 90°, was calculated on the basis of the Mie theory. The number concentration was determined from the Rayleigh ratio using the working standard method. The measurements were rapidly performed, without change of particle size during measuring time, with a device for dilution. The geometric mean diameter, the logarithmic standard deviation and the number concentration of mainstream smoke were found to be about 0.18 um, 0.4 and 3 X 10¹⁰ particles per cubic centimeter, respectively. Although the particle size presented here is smaller and the number concentration is higher than many of the previous results, the particle size is in good agreement with a few results which are considered to be more accurate. Further, the geometric mean diameter and logarithmic standard deviation of sidestream smoke .were found to be about 0.1 um and 0.4, respectively. Some previous papers have indicated that alteration of smoking conditions, tobacco type, etc. had no detectable effects on particle size; however, present work revealed a significant change in particle size and number concentration. The rapid measurement of diluted smoke by our new method seems to detect the particle size dependence on smoking variables in detail.

An important characteristic of the cut smoking tobacco is the length of the fibres. The manufacturer strives for a large mean fibre length in order to prevent the content of short tobacco from becoming too great in the course of further processing. It is important that the quality of the leaf be checked, and mainly the mean area of the strips, because this parameter is closely related to fibre length. The suitability of the leaf for the production of a cut tobacco with long fibres is designated by the so-called QV (quality relationship); the fibre distribution is characterized by the SI (sieve index). Investigations were carried out with tobacco of various provenances. The investigation and calculation methods used are described, the results given and discussed. The result shows a very good correlation between the QV and the SI. Also evident is a dependence of the content of unacceptable stem particles in the cut tobacco (URt) on the content of the stem particles "R'' (diameter more than 2.5 mm) in the threshed leaf.

The methods currently used for the determination of pesticide residues on tobacco are compared to one another, and a new method is described which results from a combination of extraction and clean-up of the extract. For that purpose, a modified Soxhlet extractor is used which serves at the same time as clean-up column. The comparison to other methods shows clear advantages as regards saving of analysis time and consumption of solvents.

Tobacco plants were separately treated with 0.75 lb./acre and 1.50 lbs./acre of endosulfan I, endosulfan II and endosulfan sulfate. Leaves from the tobacco plants were analysed for endosulfan and its metabolites at the harvesting stage, and later, when flue-cured. From the residue data obtained we conclude that [1] In green tobacco leaf endosulfan I, found in very small amounts, is the only metabolite of endosulfan sulfate, and that endosulfan II is the main endosulfan sulfate metabolite in flue-cured tobacco leaf; [2] The loss of residues during the curing process depends largely on the residue level of this pesticide in the green tobacco leaf: [3] The loss of pesticide during the curing process is mainly due to volatilization; and [4] Smallest amounts of residues are left in tobacco Ieaf with endosulfan I treatments.

It is correct to define as "condensate'' those vaporous organic products of reaction which arise in addition to the products of combustion carbon dioxide and water, when cigarettes are smoked, and which can condense when cooling and forming an aerosol of a liquid phase in a gaseous phase. This condensate is clearly defined in DIN 10 240 as "dry smoke condensate''. Reference to this composite substance as "tar'' must be rejected as it is incorrect and can lead to inaccurate assumptions regarding occurrence and composition.