Diffusion of liquids in polymers

• Published in: International materials reviews Vol. 53; no. 5; pp. 299 - 315
• Main Author: Vesely, D.
• Format: Journal Article
• Language: English
• Published: London, England Taylor & Francis 01.09.2008; SAGE Publications; Maney
• Subjects: Applied sciences; DIFFUSION; Exact sciences and technology; LIQUIDS; MASS TRANSPORT; Metals. Metallurgy; POLYMERS; POLYMERS; DIFFUSION; MASS TRANSPORT; LIQUIDS; Diffusion; Polymer
• ISSN: 0950-6608; 1743-2804
• DOI: 10.1179/174328008X324602

The purpose of this brief review is not to provide a comprehensive cover of the field, but only to open a discussion on some important issues. Many well established procedures are rarely challenged, and it may be useful to revisit them in the light of new experimental evidence. The review deals specifically with liquid transport into and through polymer matrices. It attempts to point out the accuracy of experimental observations and the ability of different theories to explain them. Most often the rates of diffusion (or time dependence of diffusion distance) are measured. The results are usually interpreted using Fick's law. This gives a good correlation for thick samples; however, for thin samples, the Case II theory provides a better correlation with experimental data. The combination of these two mechanisms, or the dependence of the diffusion distance on time to an exponent different than 1 or 1/2, however, requires inhomogeneous material. The best correlation can be obtained by considering the finite initial diffusion rate and resistance of the matrix to the flow of the penetrant (or slowing down with diffusion distance). Accurate knowledge of the concentration profile is crucial to differentiate between different mechanisms; however, no reliable analytical technique with good resolution is available. This opens the possibility to interpret the data in different ways, of which a non-Fickian interpretation is most common. It is shown that a simple reaction kinetic scheme can interpret the concentration profile data very well. It is generally accepted that the concentration gradient is the driving force for diffusion. It is, however, observed that the diffusion front is often followed by a constant concentration. In addition, voids and cavities are filled by the penetrant, representing diffusion from low to high concentration. The only models that can explain this behaviour are those considering molecular interactions. Molecular modelling of the diffusion process is improving fast with the availability of more advanced computer techniques. Simulation of the dynamic, non-equilibrium process is, however, not easy and many attempts simply lead to the Gaussian model of random motion of molecules.