Methods for Determining the Degree of Polymerization of Flavans

Abstract

Freudenberg¹, Hergert² and Hathway³ have each suggested several ways in which catechins and related compounds might polymerize to yield condensed tannins. Of these, oxidative condensation³, either with or without enzyme catalysis, appears to be the most likely way in which polymerization occurs in vivo; for example, in plum leaves during the course of the growing season⁴, or in fruits on ripening⁵. Leuco-cyanidin (3,4,5,7,3′,4′-hexa-hydroxy-flavan) which is present in plum leaves and certain fruits and is the most widely occurring flavan in plant tissues⁶ should, on Hathway's hypothesis³, polymerize by a series of ‘head to tail’ condensations forming C–C bonds between C_2′ of one molecule and C₈⁷ (or C₆) of another. As the size of such a polymer increases it is obvious that the total number of phenolic hydroxyl groups will remain the same, and therefore their reactivity with analytical reagents should be roughly constant. Since, however, the most reactive positions of the phloroglucinol (A) ring of the flavan is involved, the reactivity of the polymers to analytical reagents which only substitute in this ring should be much reduced. The same reasoning of course applies to acid-catalysed polymerizations.