3. Effect of Long Chain Concentration in Bidisperse Blends .......................III-1
3.3 Results
3.3.4 Ex situ microscopy (OM and TEM)
Optical micrographs (OM) under crossed polarizers and TEM for a B2 sample sheared at 0.055 MPa for 7.5 s (wex = 98 mg) and then crystallized at Tc = 137 °C for 2800 s are presented to show the typical morphology of blends sheared at σw < σ* (Figure 3.15). The
III-13 value of I⊥/Itot just before removing the sample was 0.075 ± 0.03. A thin (5 µm thick), somewhat oriented skin layer can be seen at the surfaces of the sample, and a fine-grained layer of spherulitic structures is observed below the skin layer. The thin birefringent skin layer is composed of stacked lamellae grown perpendicular to the flow direction (Figure 3.15B).
A similar thin layer of lamellae perpendicular to flow also appears in some of the OM and TEM images for samples sheared at σw > σ* (Figure 3.16). Previously, Kumaraswamy encountered such a layer but was unable to determine its nature.36 The replica technique used to image the features of the sample by TEM was not suitable to capture features in this dense layer. However, ruthenium tetroxide staining provides adequate contrast to distinguish the crystalline features. In the example shown (Figure 3.16), the perpendicular lamellae occur in a layer extending 6 µm from the cell wall adjacent to the highly birefringent layer and appear slightly darker in optical micrographs. At higher magnification (TEM), the morphology of the 6 µm layer appears distinctly different than the shish-kebab structures (highly birefringent layer) farther from the wall. At the highest magnification, we discern individual lamellae and can visualize the cross-hatching behavior. The high number of cross-hatches in this layer accounts for the dark appearance when viewed between crossed polarizers since the nearly perpendicular parent and daughter lamellae cancel each other’s birefringence. We believe the presence of this thin stacked lamellar layer is due to a flow effect at the wall surface (quartz windows).
Crystallization of the polymer on a quartz window substrate at the quiescent condition did not produce ordered stacks of lamellae (Figure 3.17).
All blends crystallized after shearing at σw of 0.12 MPa (> σ*) showed a highly oriented skin layer (Figure 3.18A-E). The skin layers appear as bright bands at the walls when viewed through crossed polarizers. The thickness of the skin layers for B0, B025, B05, B1, and B2 (Figure 3.18A-E) are estimated to be 10 ± 3, 25 ± 10, 53 ± 5, 47 ± 8, and 47 ± 5 µm, respectively. The thickness and uniformity of the skin layer increases remarkably at c/c* of 0.5 and saturates for c/c* =1 or higher. The position of the abrupt transition between the skin layer and the spherulitic core provides a measure of the threshold stress required to
III-14 induce the transition to oriented growth, σ*, since the stress decreases linearly from the wall to the center of the flow channel. At depths greater than 50 µm from the wall, highly oriented crystallites are not observed; the boundary is fairly sharp and corresponds to σ* ≈ 0.11 MPa. The saturation in skin thickness as a function of long chain content indicates that the threshold stress varies weakly with c for 0 < c < 0.25c*, strongly for 0.25c* < c < 0.5c*, and weakly for 0.5c* < c (Figure 3.8). As previously reported,35 the formation of the oriented skin observed ex situ correlates with the development of strong birefringence in situ after cessation of shear as lamellae grow transverse to the precursors created during shear (Figures 3.10 and 3.18).
A difference in the exact depth of the channel of flow cells containing beryllium windows used in rheo-WAXD experiments compared with the cells used for rheo-optical experiments resulted in a lower σw when the same pressure drop was applied across the cell. Therefore, the rheo-WAXD experiments required longer shearing times to apply the same total strain. Because the stress applied was closer to σ*, a thinner oriented skin layer was observed for rheo-WAXD than for rheo-optical experiments (Figure 3.19). The resulting micrographs of the quenched samples show a transcrystalline layer where the spherulites at the edge of the fine-grained layer grew toward the center of the channel. The boundary between the fine-grained and transcrystalline layer (at ~100 µm) allows us to visualize the depth dependence of shear-induced nuclei and infer the critical stress (~0.09 MPa) for pointlike nucleation induced by shear.39 (This boundary cannot be seen if samples are quenched prior to the growth of the transcrystalline layer as in Figures 3.15, 3.16, and 3.18).
TEM images of the skin layer of the same samples examined by OM reveal shish-kebab structures (Figure 3.20). The thickness of the bright skin layer observed by OM is consistent with the distance from the surface to the boundary between the row-nucleated region and the spherulitic region seen in TEM images. From a series of TEM images (Figure 3.20A-E) it can be clearly seen that the number of shish-kebabs increases with increasing c and increasing ts (Figure 3.20D and 3.20F). Thus, evidence strongly supports the notion that the strong, non-linear effect of c on the transient birefringence during shear-
III-15 induced crystallization at σw > σ* is caused by the formation of the shish-kebab structure and the non-linear increase in the number of shish-kebabs, particularly as c/c*
increases from 0.25 to 0.5.
We quantify the length of threadlike precursors per unit volume by measuring the average distance between shish-kebab centers. We determined this thread density from TEM images by counting the number of shish-kebabs in a 2 µm depth range at several distances from the wall (Figure 3.21). The inverse of the square of the distance between centers scales as shish length/vol. The thread density saturates at long chain concentration above c*.