Modeling headspace solid-phase microextraction of volatile organic compounds from water samples with porous coatings using finite element analysis

Abstract

• Model was developed for simulation of HSSPME of VOCs from water by Car/PDMS at 25 °C. • Diffusivity in fluids was estimated by Fuller, and Wilke and Chang methods. • Fiber-air distribution constants were estimated using LSER model. • Southworth approach was accurate to model mass transport from water to headspace. • The model was applied for optimizing extraction parameters and preincubation time. This research was aimed at the development of the COMSOL Multiphysics® (CMP) model for simulating the headspace (HS) solid-phase microextraction (SPME) of volatile organic compounds (VOCs) from water using Carboxen/polydimethylsiloxane coating at 25 °C. The developed model is mainly based on existing theory and previous research on the numerical modeling of SPME. Fuller method was used to estimate diffusion coefficients in air as well as pores and voids of the coating. Coating-headspace distribution constants were estimated using linear solvation energy relationship (LSER) model and multiple regression obtained by Prikryl and Sevcik. Headspace-water constants were estimated using vapor pressures and activity coefficients were determined using UNIFAC model. Wilke and Chang method was chosen for estimating diffusion coefficients in water without stirring. Liss and Slater, and Southworth approaches were tested for estimating mass transfer coefficients at the headspace-water boundary under stirring. Southworth approach allowed obtaining benzene, toluene, ethylbenzene and o -xylene (BTEX) extraction profiles from water, which were closest to experimental profiles compared to other approaches. When using Southworth approach, root mean square deviation (RMSD) between experimental and simulated values for BTEX were 8.7–10% indicating the high accuracy of the model. The developed model was successfully applied for computational optimization of extraction parameters (stirring speed, fiber insertion depth, pressure, sample volume and the concentration of added salt). After minor modification, the model was also applied for optimization of preincubation time. It can be recommended for optimization of HSSPME-based analytical methods for VOCs quantification in water.