This work describes a setup to characterize the effective diffusivity of barrier membranes to mass transport. Membranes with controlled and well‐defined barrier microstructures are produced from an elastomeric polydimethylsiloxane (PDMS) matrix. Physical obstacles to mass transport are generated by carving in the matrix microscopic holes with well‐defined shapes and in specific positions. Lateral Raman mapping is used to follow mass transport of a penetrant molecule, hexadecanol (HDOL), in the carved matrix. The influence of circular and elongated holes over mass transport is examined by monitoring the HDOL Raman intensity over time, after it passes the barrier microstructure. Results are compared with computer simulations of mass transport and excellent agreement is found. It is shown that holes act as bidimensional barriers to HDOL and that the effective diffusion coefficient of HDOL in the barrier membrane is reduced with respect to that of the homogenous material. Overall, Raman measurements provide a valuable base of data generation to contrast theoretical models of mass transport in heterogeneous barrier membranes.

This paper compares direct calculations, experimental data and analytical models in materials with controlled barrier microstructure. Confocal Raman was used to follow mass transport of a penetrant, hexadecanol (HDOL), through the barrier structure. produced by carving holes in specific positions and orientations in a polydimethylsiloxane (PDMS) matrix. Finite element method (FEM) was used to predict maps of HDOL concentration around the obstacles, which were directly compared to Raman data. Effective diffusion coefficients were obtained in homogeneized microstructures of slender obstacles in diluted and semi‐concentrated obstacles regimes. Results were compared with predictions of simple analytical models: Nielsen, Lape, those derived from the ideas of Bharadwaj (Greco; Greco and Maffezzolli), and those developed from FEM simulations (Minelli et al. and Dondero et al.). For the first time, predictions of these widely used analytical models could be tested against well‐controlled barrier membranes. The influence of obstacle orientation, size polydispersity and eventual aggregation, all features occurring in real polymeric nanocomposites are analyzed. Some limitations about the use of simple analytical models to interpret permeability data in real nanocomposites are discussed.