Column fractionation of polymers

Abstract

AbstractMixtures of fractions of atactic polystyrene dissolved in cyclohexane were resolved by elution from a column at different temperatures. A coarse‐grained Ottawa sand was used to minimize the surface area of the support. Comparison with the critical miscibility temperature data of Schultz and Flory shows that the lower molecular weights are eluted from the column at a temperature somewhat below the critical miscibility temperature, as expected, but for molecular weights above 3 × 105 the elution temperature increases abruptly, and even exceeds the critical miscibility temperature. This rapid rise occurred at 1.5 × 105 when the surface area of the support was increased by use of a very fine Celite. We conclude that for the higher molecular weights additional thermal energy is required to desorb the polymer, and hence the partitioning of these species is not determined solely by solubility considerations. Concerning the efficiency of the column extraction, considerable resolution of each component was obtained, and attempts to determine the M̄w/M̄n ratios of the individual components were at least partially successful. As predicted by theory, for a given molecular weight interval the range of elution temperature was increased fivefold by the use of a mixed solvent, benzene–ethanol, in place of the single poor solvent, cyclohexane. Customary fractionation procedures are very inefficient when applied to high‐melting crystalline polymers. For these polymers we investigated selective precipitation upon a column having a temperature gradient, rather than elution from the column. For dilute solutions, amorphous phase separation may precede crystallization, particularly in the case of polymers whose rates of crystallization are low, owing to the presence of bulky side groups. The column precipitation method failed to give improved fractionation of polyacrylonitrile, owing to phase separation of the dimethylformamide–heptane mixture part way down the column. On the other hand, good resolution was obtained for an unfractionated sample of isotactic polystyrene up to 1.5 × 106. Molecular weights above this level were compressed into a narrow band at the top of the column. Thus, if the distribution is extremely broad, a second fractionation with a different solvent:‐ precipitant ratio may be required. Even so, a single pass through the column resulted in a better resolution than that obtained by a large‐scale, two‐stage extraction‐precipitation procedure, indicating that the column precipitation method does posses an advantage for the fractionation of crystalline polymers.