Functionalizing polyethersulfone membranes: using NMR to avoid pitfalls when using UV-induced polymerization to 'graft from' surfaces

For nearly 30 years, UV-induced free radical polymerization has been used to modify the surface chemistry of polyethersulfone (PES) membranes, films, and coatings. The initial mechanism for the grafting process was introduced in the 1990s and supported without direct evidence of covalent bond formation. Since then, claims of ‘grafting from’ membranes in the literature rely on similar evidence using a combination of gravimetry, attenuated total reflectance infrared spectroscopy, contact angle goniometry, water sorption, and/or water permeance. While these techniques provide evidence of the presence of a polymer coating, they do not provide direct evidence of covalent bond formation between the grafted polymer and the PES membrane or support. Thus, it is possible to mischaracterize a physisorbed film as a chemically bonded grafted polymer. We use experimental data from a model system to examine two key areas: 1) the impact of experimental parameters (UV intensity, UV wavelength, grafting time, and monomer concentration) on the degree of grafting and 2) the limitations of the current literature-accepted techniques used to characterize surfaces functionalized by UV-grafting. NMR is introduced as a technique to quantitatively analyze grafting from a PES membrane and is compared directly to the literature-accepted techniques to expose ‘false positives’ that may lead to incorrect conclusions about ‘grafting from’. Finally, we use 2D NMR techniques to present direct evidence of covalent bond formation from the UV-initiated reaction between a model PES monomer and a model methacrylate monomer in solution.