Heats of polymer mixing

Abstract

Properties of compatible multicomponent polymer systems can be studied quantitatively only if their fundamental thermodynamic parameters can be established. Thus measurements of mixing enthalpies in such polymer-polymer pairs is of interest in arriving at criteria for compatibility. Direct determinations are awkward, but the required parameters can be obtained from differential heat of solution measurements. The metastability of polymer blends in phase separated regions below their glass transition temperatures provides additional insight when calorimetric measures are performed below T . These factors are illustrated with examples of compatible, near-compatible and incompatible binary mixtures containing poly(2,6-dimethyl phenylene oxide) and polystyrene or halogenated derivatives of polystyrene. INTRODUCTION A quantitative understanding of the properties of compatible polymer systems polymer blends requires a determination of phase boundaries and of the thermodynamics (and kinetics) of the mixing process. This understanding has come about only quite recently and only in a few cases. The difficulties encountered include (1) the comparative rarity of the occurrence of miscible binary polymer systems, (2) the difficulty of establishing thermodynamic equilibrium at temperatures approaching the glass transition of the blend, and (3) the sensitivity of compatibility to effects of molecular weight (and presumably molecular weight distribution) and to changes in chemical structures of either or both of the prospective constituents, (1,2). Several polymer-polymer blends are known to be compatible over only limited regions of the temperature-composition plane, (3). A two phase, incompatible, regime may be present at all temperatures for a given composition (usually around equal volume fractions of the constituents)or there may occur a lower critical solution temperature (LCST). Because of the difficulty in attaining equilibrium already referred to it is normally necessary for the LCST to be located some tens of degrees above the T for a definitive determination. It is also possible that a second, upper consolute point (a UCST) is present in certain cases, though again direct observation of the critical temperature and associated phase boundary lying at or below T , is not possible, (4). The research reported below describes a technique for overcomin this difficulty; in several systems evidence for the existence of a UCST has been obtained. The upper portion of Fig. 1 illustrates schematically a typical situation. It should be noted that the comparatively slow kinetics of phase separation and rehomogenization can be taken advantage of in the determination of phase boundaries in polymer-polymer systems. In establishing such a boundary it is usually not practical or desirable to attain the macroscopic phase separation or layering typically observed in low molecular systems. Instead the determination of either one or two phase states for a particular composition-temperature locus is correlated with the observation of, respectively, a single or double glass transition. It is clear that this criterion may be somewhat compromised by a lack of understanding of the relation of domain size to glass transition manifestations, (5). The determination of the phase boundary or boundaries represents only a portion of the thermodynamic information required to characterize a polymer blend. A complete definition requires a knowledge of the free energy of mixing tG as a function of composition and temperature (from which data, of course, the phase bundary may be obtained). For binary polymer systems where each constituent is of high molecular weight it is anticipated that the entropic contribution tSm to tGm will be orders of magnitude smaller than in low molecular weight mixtures (though non-configurational contributions to AS itself may be significant) and that the enthalpic term Hm will therefore be dominant, (1). Experimental determinations of Hm for polymer mixtures have been rare, reflecting the difficulty of such measurements. Direct measurement would appear to be unfeasible; two indirect approaches have been attempted. In the first, tHm for low molecular analogies (or if possible oligomers) of the respective polymers have been measured by conventional means