Why do polyethylene crystals have sectors?

High-resolution (3.7 A in optical diffraction) electron microscope images have been obtained from a series of n-paraffin monolamellar crystals with chain lengths from n-C36H74 to n-C82H166. The higher molecular weight specimens, which do not undergo chain folding, form sectorized crystals and the molecular packing is found to include alternate bands of untilted and tilted chains along <130>. Their widths are consistent with those of Bragg fringe widths in bright-field images obtained at lower magnification. The chain tilt axis is near d*110. Lower molecular weight paraffins form nonsectorized crystals where the chains are generally untilted with occasional small inclinations around nonspecific axes. Surface decoration of the longer alkanes with polyethylene crystallites, first of all, reveals three preferred polyethylene crystal rod orientations ([100] plus two perpendicular to [110]) instead of the two commonly found for the lower alkane. Control studies on solid-solution crystals reveal that the third [100] orientation is a result of slight surface roughness due to unequal chain lengths or surface protrusions of chains; the new decoration is also randomly distributed. For pure n-C60H122 lamellae, however, suggestions of regular bands containing rods along [100], due to surface discontinuities along <130>, can also be seen. In contrast with polyethylene, these data suggest that crystal sectorization may be a function of chain-stem packing alone and that chain folds may play merely a secondary role in the polymer--e.g., by directing the collapse of pyramidal crystals on a flat surface.