Extracting physically useful information from multiple-detection size-separation data for starch

A method for interpreting multiple-detection size separation data of complex branched homopolymers [Konkolewicz, D.; Gilbert, R. G.; Gray-Weale, A. Phys. Rev. Lett. 2007, 98, 238301] is applied to starch. The method, whose application is described in detail, uses the sample's weight and number distributions over polymer sizes, along with the molecular weight distribution of the individual branches (or their average degree of polymerization). The branch-length and number size distributions are used to generate the weight distribution of a hypothetical molecule with the same branch-length and number distributions but where the branches are randomly joined; this reference weight distribution is then compared to the actual one. The method is applied to size-exclusion chromatography (SEC) data for starch from a particular rice variety, the first time such data have been reported for a native starch. Comparison with the randomly branched reference function shows that the amylopectin component is consistent with random branching on the distance scale of this measurement, 10(2)-10(3) nm. This implies that on the size scale commensurate with that of a whole amylopectin chain, branching is pseudorandom, even though there is nonrandom branching on the much smaller scale of individual branches and clusters.