Estimating Air Concentrations

Levels of organic pollutants in the atmosphere are subject to variability. For many semivolatile organics this variability in air concentration has been shown to be correlated with changes in emission source strength, air temperature and other meteorological factors. For instance, rain or snow can effectively scavenge semivolatiles from the atmosphere, lowering their air concentrations. Air masses that have moved over source areas before arriving at the sampling site can accumulate the contaminant in question, causing concentrations to become elevated. Consequently, depending on deployment duration, active air sampling may provide only a 'snap shot' of air concentrations at a given site. Passive air samplers allow a time-integrated air concentration to be determined. On this basis, active samplers may be useful in assessing how concentrations of a substance may change in response to changing meteorological conditions or emission source strength, whereas the passive air sampler is well suited for monitoring changes in concentration from one period to another. The latter is useful in assessing whether or not levels of banned substances are declining as a result of regulatory activity.

In order to use passive air samplers to measure atmospheric concentrations of pollutants, calibration data are required. Calibration data include parameters such as sampling rates, sampler/air partition coefficients and loss rate constants [23]. These parameters are usually determined in the laboratory, at a reference site or in situ.

When using passive sampler data to estimate the air concentration of SOCs, investigators commonly assume that sampling follows first-order exchange kinetics. Thus, during the first stage of sampler uptake, chemicals are accumulated linearly relative to time. Later, as the outward flux of accumulated chemicals slowly increases, sampling moves into a curvilinear stage and eventually, sampling reaches an equilibrium stage where analyte uptake and loss fluxes are balanced (Fig. 6.1). Because the physicochemical properties of SOCs vary widely, the estimation of atmospheric concentrations from passive sampling data may require the use of all stages of this model. For example, if uptake remains in the linear phase (i.e. loss from the sampler is negligible) during an exposure, then the sampling rates that describe the extraction of chemicals from a certain volume of air over time (i.e. metre cube per day) are used along with the amount of pollutant quantified in the sampler to estimate air concentrations. Before quantitative determinations of atmospheric chemical concentrations from passive samplers

Fig. 6.1. Uptake curve for a passive air sampler showing the linear, curvilinear and equilibrium stages of analyte accumulation [23]. Reprinted from Ref. [23]. Copyright (2005), with permission from Elsevier.

Curvilinear stage

Approach to equilibrium

Time are possible, investigators must have a high level of confidence that the model used for concentration extrapolations fits sampler exchange kinetics and that the calibration data used closely reflect actual in situ sampling rates.

A range of studies have made comparisons between air concentrations measured using passive and active samplers. Shen et al. [22] have used XAD-based passive air samplers to measure the air concentration of organochlorine pesticides throughout North America. These values compared reasonably well with recent measurements made using active sampling equipment in other studies. Work reporting PBDE and ^PCB air concentrations in Canada using PUF disks and active samplers have also shown good agreement [16,24]. Van Drooge et al. [25] reported good agreement between SPMDs and active samplers when measuring the air concentration of PCBs and HCB at sites in the Pyrenees, with many values comparing within a factor of 2. Jaward et al. [26] used PUF and active samplers to measure the air concentration of PCBs and HCB in Italian mountain air, also reporting that agreement was generally within a factor of 2. Further studies from North America have reported comparable measurements using PUF and active sampling data for hexachlorocyclohexane [27] (Fig. 6.2).

1 10 100 1000 Active sampler based air concentration (pg m 3)

Fig. 6.2. Plot of alpha-HCH air concentrations measured using active and passive samplers at sites in Canada. The error bars indicate the maximum and minimum measurements made by the active samplers. The dashed lines represent where the two measures of air concentration vary by a factor of 2.

1 10 100 1000 Active sampler based air concentration (pg m 3)

Fig. 6.2. Plot of alpha-HCH air concentrations measured using active and passive samplers at sites in Canada. The error bars indicate the maximum and minimum measurements made by the active samplers. The dashed lines represent where the two measures of air concentration vary by a factor of 2.

Bartkow et al. [28] showed that SPMDs could be used reproducibly to estimate air concentrations of various PAHs at an urban site in Brisbane, Australia. The difference between SPMD-derived air concentrations and those measured using active samplers was within a factor of 2. Jaward et al. [29] achieved similar results using PUF disks to estimate the air concentration of PAHs at four sites in the UK. The outliers reported by Jaward et al. [29] were from one site, and all other compared values were within a factor of 2. Both studies reported good estimates of PAH concentrations for the higher molecular weight compounds, which are predominantly associated with particles (Fig. 6.3).

The good agreement between passive sampler-based air concentrations and measurements made using active samplers for PAHs predominantly associated with particles is interesting. Models used to describe chemical exchange by passive samplers were developed to describe the uptake of vapour phase compounds which accumulate in the sampler by diffusion from the air via first-order kinetics. At this stage, it is unclear which processes dominate the uptake of predominantly particle-bound SOCs.

Bartkow et al. [28] suggested that the dominant uptake process for particle-bound PAHs (i.e. those with log n-octanol-air partition coefficients (KOAs) > 9) sampled by SPMDs was via particle deposition. These

0 0

Post a comment