Determination of time-weighted average (TWA) concentrations of volatile organic compounds (VOCs) in air using solid-phase microextraction (SPME) is advantageous over other sampling techniques, but is often characterized by insufficient accuracies, particularly at longer sampling times. Experimental investigation of this issue and disclosing the origin of the problem is problematic and often not practically feasible due to high uncertainties. This research is aimed at developing the model of the TWA extraction process and optimization of TWA air sampling by SPME using finite element analysis software (COMSOL Multiphysics, Burlington, MA, USA). It was established that sampling by porous SPME coatings with high affinity to analytes is affected by slow diffusion of analytes inside the coating, an increase of their concentrations in the air near the fiber tip due to equilibration, and eventual lower sampling rate. The increase of a fiber retraction depth (Z) resulted in better recoveries. Sampling of studied VOCs using 23 ga Carboxen/polydimethylsiloxane (Car/PDMS) assembly at maximum possible Z (40 mm) was proven to provide more accurate results. Alternative sampling configuration based on 78.5 × 0.75 mm internal diameter SPME liner was proven to provide similar accuracy at improved detection limits. Its modification with the decreased internal diameter from the sampling side should provide even better recoveries. The results obtained can be used to develop a more accurate analytical method for determination of TWA concentrations of VOCs in air using SPME. The developed model can be used to simulate sampling of other environments (process gases, water) by retracted SPME fibers.
Keywords: COMSOL; air analysis; air sampling; solid-phase microextraction; time-weighted average; volatile organic compounds.