The Molecular Basis of Resin Behavior.-E. 0. KRAEMER. (Condensation of paper presented at Gibson Island Conference on Organic High Molecular Weight Compounds, June 8, 1940. (Continued from page 518.)) Non-Uniformity of Molecular Size.
With the exception of some crystallizable proteins and certain synthetic polymers prepared by quite special methods (for instance Hibbert's polyoxyethylene glycols) all soluble macromolecular materials are mixtures, and the complete definition of molecular weight requires the use of a size-distribution curve.
Non-uniformity is most readily demonstrated by fractional dissolution or precipitation and comparison of the properties of the fractions.
Clean-cut separation of the different molecular species is of course not possible, and a reasonably accurate sizedistribution curve requires repeated fractionations. Spurlin probably holds the record for extensive fractionation with his separation of a low-viscosity nitrocellulose into 67 fractions. Schulz has published size-distribution curves for polystyrenes involving 10-12 fractions, using them to confirm'theoretically calculated distributions derived from the kinetic analysis of the process of polymerization and condensation (see papers by Flory, Mark, Schulz and others).
The most generally applicable direct method for determining heterogeneity is by means of the sedimentation equilibrium with the ultracentrifuge.
Unfortunately, it does not give a detailed representation of the heterogeneity, but rather an equivalent heterogeneity which may be expressed in terms of a suitable empirical distribution function.
Consideration of the interpretation of the sedimentation equilibrium of mixtures led to a method of characterizing heterogeneity in terms of what we called weight-average and Z-average molecular weights, both of which can be calculated from the ultracentrifuge data, provided all molecular species have the same V-value and light absorption or refractive increment at some