Non-university of Constraint Release Relaxation in Entangled Linear Polymers of Various Chemical Structures

Abstract Extensive experiments have established that the constraint release (CR) relaxation takes place in binary blends of chemically identical long and short polymer chains wherein the long chains are dilute and entangled only with the short chains. Recently, Hannecart and coworkers focused on polymers of various chemical structures, polystyrene (PS), polyisoprene (PI), polybutadiene (PB), and poly(methyl methacrylate) (PMMA), and compared the CR relaxation time of the long chains in the binary blends of each polymer species (Hannecart, Clasen, and van Ruymbeke, Polymers 15, 1569 (2023)). From this comparison, they concluded that a normalized lifetime of the entanglement obstacle, with ZL = ML/Me (entanglement number of the long chain) and τe being the Rouse relaxation time of the entanglement segment, is determined only by the entanglements number of the short chain, ZS = MS/Me, irrespective of the chemical structure of the chains. This universality (independence from chemistry) would be an important feature if it were unequivocally concluded from experimental data. However, the values of the molecular weights utilized in their comparison, ML, MS, and Me, should have unavoidably included experimental uncertainties, which disturbs rigid conclusion of the universality. Aiming at a rigid experimental test avoiding those uncertainties, this study focuses just on data of the linear viscoelastic moduli G* of entangled monodisperse polymers of various chemical structures, PS, PI, PMMA, and poly(t-butyl styrene) (PtBS). We were able to find several pairs of chemically different but viscoelastically equivalent monodisperse polymers exhibiting indistinguishable G*GN data (with GN being the plateau modulus) from the local Rouse relaxation zone to the terminal relaxation zone. For binary blends of those equivalent polymers in each chemical species, i.e., long-X/short-X blends with X = PS, PI, PMMA, or PtBS, our experiments revealed that the CR relaxation of the dilute long chain does not complete at the same reduced frequency wte even when the chemically different component chains were viscoelastically equivalent in their monodisperse bulk state. It turned out that the CR relaxation is slower in the order of PS (slowest) < PMMA < PI < PtBS (fastest) and this difference was by a factor of 3-4 in total (well above the experimental resolution limit), rigidly showing the non-universal character of CR. An origin of this non-universality is briefly discussed within the framework of existing CR models, for example, Graessley’s CR model that already involved a chemistry-dependent parameter z representing a number of local CR hopping sites per entanglement segment.