On quantitative permanganic etching

Permanganic etching exposes lamellae within a variety of crystalline polymers. It is frequently observed that certain lamellar populations and other regions are removed preferentially by the etchant. This selectivity has been studied quantitatively by using 5 μm sections of a linear polyethylene that had been crystallized to form two populations, one of which contained lamellae roughly half as thick and molecules half as long as the other. Changes in the melting endotherm of sections with the time of etching and the related mass losses have been studied in relation to electron microscopy and molecular-mass data of polymer extracted from the sample with xylene. It was found that in sections cut at 20°C, the lower melting population suffered substantially greater deformation, but this could be limited by cutting at lower temperatures. The changes undergone by cold-cut sections were linear with etching time and revealed preferential removal of the thinner lamellae at a rate of 3.3 Å s −1, compared to 1.6 Å s −1 for the other population. It was also found that if sections were annealed at a temperature between the two melting peaks, a treatment giving similar populations to the original but with sectioning damage healed, then linear etching continued at a rate of 1.7 Å s −1, but now with little or no discrimination between the two populations. It is concluded that when differential permanganic etching is observed it is not necessarily, at least in this case, an intrinsic property related to lamellar thickness or molecular mass, but reflects a secondary effect, namely the different responses of elements of physical texture to stresses imposed during sample preparation. Conversely these findings illustrate very well that the systematic spatial variation of texture imposed on a polymer by crystallization results in a corresponding systematic and local variation of properties.