Air sampling technologies: principles and applications. Cincinnati, OH: American Conference of Governmental Industrial Hygienists, 2013 Aug; :1-22
Particle and gas phase interactions can lead to errors in air sampling results. Errors may result from allowing unwanted chemical reactions to occur during sampling, measuring the concentration of a contaminant only in the particle or gas phase when there is a significant fraction in the unsampled phase, and using inappropriate techniques to measure the distribution of a contaminant between the two phases. The distribution of an air contaminant between the particle and gas phases (and the potential for changes in that distribution) have important implications for air sampling strategies. Three pieces of information are helpful in judging whether it is necessary to sample the gas, particle, or both phases of an atmosphere: the saturated vapor concentration of the contaminant, a rough estimate of the total airborne concentration of the contaminant in the atmosphere being sampled, and a rough estimate of the concentration and composition of the particles in the atmosphere. The distribution of a contaminant between the particle and gas phases depends on temperature, pressure, particle composition, and the gas phase concentration of the contaminant. The processes of sampling or inhalation may change the distribution between the phases significantly. Knowing the total airborne concentration of a contaminant, i.e., the concentration in the particle phase plus that in the gas phase, is perhaps more important than measuring the distribution between the phases. The total airborne concentration can be measured by a sampler consisting of an efficient filter followed by an efficient sorbent or can be measured by a filter that is treated with a material that adsorbs or reacts with the vapor. There is no universally acceptable technique for measuring the distribution of a contaminant between the particle and gas phases. The common approach of placing a sorbent downstream of a filter is prone to errors if the filter adsorbs vapor or if the atmosphere's composition changes. Replacing the filter with a virtual impactor reduces many of these errors. The approach of placing a denuder capable of collecting the vapor upstream of a collector of the particles and vapor that penetrate the denuder is prone to errors if the denuder is not optimized for the application. In some cases, it may be difficult to avoid significant errors due to evaporation of the contaminant from the particles while they reside in vapor-depleted air in the denuder. A third approach to measuring the distribution of a contaminant between the gas and particle phases is available. The total airborne concentration is measured by an active sampler and the vapor concentration is measured by a passive sampler. This approach has not been widely used, yet. Other approaches are likely to be suggested in the future, but measuring the distribution will always be much more difficult than measuring the total airborne concentration.