CORESTA Congress, Shanghai, 2008, SSPT 44

High throughput puff-by-puff quantification of smoke analysis

TROUDE V.; DUPUTIÉ G.; DUVAL G.
Altadis - Imperial Tobacco Group, Research Center, Fleury les Aubrais, France

The analysis of cigarette smoke collected on a smoking machine shows that the composition and yields of smoke constituents vary from puff to puff, and depend on the smoking regime. A programme of work has been carried out to minimise the number of analyses and time required. Reducing time needed for sampling and analysis resulted in improved laboratory efficiency while using the same set of analytical methods. Taking into account the chemical properties of volatile and carbonyl compounds in smoke, two new parallel devices have been implemented to achieve high throughput puff-by-puff quantification. Firstly, an automatic puff-by-puff sampling smoking system, followed by a classical "off-line" Hoffmann determination will be presented. The application of a specific smoking robot VC10 from Vitrocell→ in the ISO vs. Intense Regime comparison of carbonyls puff by puff delivery will be shown. Special attention was given to the geometric design to provide a good day-to-day variability. The limiting factor was then the HPLC determination. Secondly, for volatile compounds, an "on-line" gas phase puff by puff quantification was achieved. The aim was to shorten the gas chromatographic cycle time, while maintaining resolution. A smoking machine coupled to a GC-MS system with Gerstel Modular Accelerated Column Heater was proposed. This Low Thermal Mass column oven was used for fast capillary GC analysis, with a target analysis time under one minute. The resulting improvement was significant: separation of mainstream smoke was achieved in approximately 0.6 minutes. Cool-down and equilibration of the column was achieved in 0.4 minutes. Comparisons with Tedlar bag and current impinger methodologies were made. The simplicity of the use of an automated valves system and flexibility for a variety of samples with consecutive run acquisitions will be demonstrated.