A Critical Assessment of Recent Work on the Application of Gas/Particle Partitioning Theories to Cigarette Smoke

In the August 1997 issue of Environmental Science&Technology (ES&T), PANKOW and co-workers at the Oregon Graduate Institute reported that the addition of ammonia-containing additives to cigarette tobacco increased the amount of unprotonated nicotine in cigarette mainstream smoke (MSS) and thus increased the bioavailability of nicotine to the smoker. Articles about PANKOW's work also appeared in other publications along with allegations that ammonia-containing additives are used to manipulate nicotine deliveries. However, initial review of PANKOW's research and that reported on environmental tobacco smoke (ETS) in an earlier paper showed that potentially serious issues existed with PANKOW's experimental data and conclusions. Consequently, a critical assessment of PANKOW's research and the underlying theories of gas/particle partitioning was undertaken. This assessment confirmed that PANKOW and his co-workers made a number of errors not only in their determinations of the gas/particle partitioning coefficients for nicotine in MSS and ETS but also in the interpretations of the data. During the preparation of this assessment, data from other researchers became public. These data showed that there was no correlation between tobacco ammonia (including residual ammonia from the use of ammonia-containing additives) and MSS ammonia deliveries and MSS smoke pH, and that the amount of unprotonated nicotine in the undiluted MSS of a full flavor (FF) American filter cigarette was less than 0.1%. These new data fully substantiated the findings of this assessment, and it can be safely concluded that the assertions made by PANKOW and his co-workers were incorrect. However, this assessment also showed that there is significant merit in the application of PANKOW's theory of absorptive partitioning for the estimation of the gas/particle partitioning of semivolatile components in MSS and ETS. Application of PANKOW's theory along with data from recent tobacco related conferences has allowed estimation of the gas/particle partition coefficient for nicotine in cigarette MSS and also has allowed approximation of values for the activity coefficient of nicotine in mainstream particulate matter.