Diffusive uptake in passive and active adsorbent sampling using thermal desorption tubes.

Abstract

Low flow active sampling techniques collecting vapors and gases using thermally desorbable adsorbents are now feasible and desirable in many applications as they permit long integration times, the potential for miniaturized sampling configurations, and other advantages. At very low air flow rates (< 1 ml min(-1)), diffusive uptake on adsorbents in conventional sorbent tubes may equal or exceed the active (pumped) uptake rate, and even at low flow rates (1-4 ml min(-1)), diffusive uptake may significantly bias measurements. Thus, corrections to account for the diffusive flux or means to limit the diffusive uptake are needed in low flow applications. This paper presents (1) a theoretical analysis of the role of diffusive and advective uptake for several sampling geometries of tube-type samplers; (2) experimental confirmation using both laboratory and field studies; (3) estimates of the tortuosity and porosity of the glass wool packing used to retain the adsorbent, parameters needed to estimate diffusive fluxes in passive and active sampling; (4) a demonstration that orifice-equipped low flow active samplers can reduce diffusive uptake and improve precision, and (5) a model predicting the saturated adsorbent layer that helps to account for the gradual decline in uptake rates seen in passive sampling. Diffusive uptake will depend on the tube configuration and diffusion coefficient of the substance of interest, but for conventional sampling tubes (0.4-0.5 cm id, 1.5 cm air gap), sample flow rates should be maintained above 1 to 4 ml min(-1) to keep errors below 5%. Laboratory experiments showed close agreement with theoretical calculations, and the field study using 1 to 4 d sampling periods and 0.3 ml min(-1) flows demonstrated that the orifice-equipped samplers essentially eliminated diffusive uptake. No significant practical difficulties are encountered using orifices, e.g., pressure drop is minimal. Experimental estimates of tortuosity (0.79 +/- 0.02) and porosity (0.92 +/- 0.10) of the glass wool packing (0.3 cm length) represent relatively little resistance to diffusion; however, variation in the packing and adsorbent placement can degrade the precision achievable by passive samplers. Diffusion barriers, consisting most simply of an orifice, may be used to lower the diffusive uptake. A needle-type orifice permits flows below 0.1 ml min(-1) and is suitable for sampling periods as long as several weeks, and it provided greater precision than conventional open-ended sampling tubes (8% compared to 13%). Finally, the gradual decrease in diffusive fluxes often seen in passive sampling is attributed to additional resistance posed by a saturated adsorbent layer, in agreement with a simple model based on total VOCs and specific adsorptivity of the adsorbent.