One of the problems in maintaining indoor comfort involves controlling the tobacco smoke odor (1). Efficiency evaluation of various approaches to reduce or eliminate this odor in occupied spaces has been hampered by lack of a simple and uniform test for all air purification devices or odor control processes. Such a test would be possible if the compounds responsible for the odor were known and tobacco odor simulant mixtures of appropriate single chemicals could be prepared. The present study constituted the first step toward finding correlates between the tobacco smoke odor and some gas-chromatographic features of the smoke.

It was considered important to base the smoke odor simulant mixture on the odorants authentically present in tobacco smoke. Only an authentic mixture would behave physically and chemically similar to the odorous tobacco smoke when treated by various odor mitigation processes.

Although literature lists many hundreds of compounds found in tobacco smoke (2,3,16), it is highly unlikely that all are odor-relevant at the concentration levels occurring in smoking rooms. It can be argued that different tobacco products produce smoke with different concentrations of various odorants, and that some components may be generated only by certain tobacco products, so that a tobacco smoke odor simulant would be of limited value. However, the odor of tobacco smoke is easily discriminated from many other occupied space odors. Hence, much of its character and intensity should be carried by a limited number of components common to most tobacco smokes. An analogy may be drawn to the findings on odors of some food substances. Even if hundreds of aromatic substances are known to occur in apple essence, Guadagni (4) has shown that only three specific compounds suffice to produce a similar aroma. Von Sydow et al (5) have demonstrated a similar simplicity for the aroma of bilberry. It was hoped that reasonably simple relations may also be found in the case of tobacco smoke odor