Sub-T g Transitions

Abstract

There have been a number of good general reviews of relaxation processes that occur at temperatures below the glass-transition temperature (T g) [1–6]. Early dynamic-mechanical studies were surveyed by Woodward and Sauer [7]. Although now outdated, the seminal work reviewing both dynamic-mechanical and dielectric data is still the excellent 1967 monograph by McCrum, Read, and Williams [8]. It is not the purpose of this review to approach the comprehensive coverage provided by McCrum et al. but to summarize important results for the major polymer groups and to include more recent studies, especially for engineering thermoplastics not avail-able prior to 1967. Where appropriate, mention is made of recent efforts at molecular modeling that aid in understand-ing the nature of molecular processes that operate below T g . A variety of techniques can be used to detect relaxational processes occurring below T g . These include dynamic-mechanical, dielectric, NMR (e.g., 1 H line width and pulsed 13 C NMR relaxation times), and thermally stimulated dis-charge current (TSC) measurements. Of these, dynamic-mechanical methods [9] have been the most widely used to study secondary-relaxation processes in polymers. These include free vibration methods, principally torsional pendu-lum [10,11] and torsional braid [12], and forced oscillation (FO) methods utilizing mechanically driven tensile, tor-sional, and flexural strains provided by a number of com-mercial instruments. The ability to vary oscillation frequency over a wide range makes FO (ca. 0.016–160 Hz) and dielectric techniques (ca. 10---10 6 Hz) especially useful for the determination of activation energies. Early dynamic methods also included resonance electrostatic methods that provide dynamic data over a higher (acoustic) frequency range (ca. 10 3 ---10 4 Hz) than is possible using FO dynamic mechanical methods. Basic principles and instrumentation for resonance electrostatic and other dynamic measurements are given by Ferry [13]. In the following sections, the results of dynamic-mech-anical and dielectric measurements of important amorphous and semicrystalline polymers are summarized and conclu-sions regarding the origin of sub-T g molecular motions are offered. As illustrated by Fig. 13.1, temperature assignments for principal relaxations are the temperatures at the max-imum of the dynamic-mechanical or dielectric tan d or loss modulus peaks at the reported frequencies. Values deter-mined from tan g data are slightly higher than those deter-mined from loss modulus values, and temperature assignments increase with increasing frequency. Typically, peak assignments for T g are slightly higher (ca. 15–20 8C) than obtained by dilatometry at low cooling rates. Where available, data for only dry, unconditioned samples are reported in this review.