3.4 Comparison of pump sampling and stir bar sorptive extraction
Because the amount of PAHs extracted with a given SBSE device from the indoor air cannot be assigned to a specific air volume in contrast to pump sampling where the analyte content is filtered out from a defined volume there is no directly obvious correlation between both approaches.
SBSE is far from reaching saturation in the time scale of several days but obviously exhibits detectable adsorption of each PAH congener already within one day. The course of further adsorption over consecutive days is obviously congener-specific and subjected to a continuous sorption/desorption dynamic triggered by temperature and air exchange. In contrast, pump sampling is unaffected by moderate air exchange due to the short sampling period and desorption of compounds once sampled on the filter is completely excluded.
Nevertheless, both sampling principles display largely similar congener patterns which means that SBSE does not discriminate between PAH congeners significantly different than pump sampling and can be regarded as equivalent in this regard. Both principles are expectedly influenced by temperature as this triggers the desorption of PAHs from the source in the first place.
It is seen that absolute absorbed masses on a SBSE device after one day in the first campaign ranged between 40 - 100 ng in case of acenaphthylene and anthracene, between 100 and 600 ng in case of naphthalene, acenaphthene, fluorene and pyrene, between 1000 and 2000 ng in case of phenanthrene and fluoranthene as the most abundant congeners, and up to 15 ng in case of benzo[a ]pyrene and the other 4 and 5 membered ring congeners (Fig. 2 and supplementary Table S4).
In comparison to that pump sampling revealed airborne concentrations in the warmer weather samplings (intraday 1 and interday, Fig. 3 and Tables S7 and S8) up to about 100 ng m-3 in case of naphthalene, acenaphthene, fluorene and pyrene, up to between 250 and 800 ng m-3 in case of phenanthrene and fluoranthene and up to about 2 ng m-3 in case of the other 4 and 5 membered ring congeners. Both sampling concepts revealed lower concentrations or absolte amounts in case of cooler weather (2nd SBSE campaign and intraday 2). This information on absolute PAH level sampled by SBSE in comparison to pump sampling should be sufficient to decide on the necessity of pump sampling providing a mass per volume estimate of PAH contamination. In addition, the dynamics of SBSE may be explored over a period of time and range of conditions to select pump sampling conditions regarded as most appropriate. It should be noted that in settings with less or controlled air exchange an even better correlation between both sampling approaches may be observed.