Domain Formation, Transformation and Destruction

As the last step in nanostructure evolution analysis we are searching for hints concerning the formation, transformation and destruction of domains. Initial strain may arrange hard segments in such a way that they form new hard domains. Further straining may break intermediate hard domains, resulting either in a transformation of thin-layer sandwiches into thick-layer sandwiches or in a complete loss that decreases the number of sandwiches. 

In order to answer this question we study meridional slices fzfr)] i (r3 = z r 2 = 0, Ti) of the CDF. It appears worth to be reminded that such slices in real space, r, describe the correlation among the phase boundaries between hard and soft domains in the direction of strain. Assuming that there is no relevant correlation beyond the intra-sandwich ones, the curve-z(ri2 = 0, r3 cut from the CDF on its meridian must be positive for r ,. r t is the maximum thickness of a single domain. Then— 

For all materials the nanostructure shows a broad variety of scattering entities ranging from thin-layer sandwiches to thick-layer sandwiches. After the start of the experiment the thin-layer-sandwich maximum is increasing. Thus the number of thin-layer sandwiches is increasing for TPU205 and TPU215 up to 0.25, and for TPU235 up to 0.5. An increase of contrast between the hard phase and the soft phase could cause the same effect, but it appears unlikely to assume that the soft-domain density decreases as a result of applied strain. The more probable formation 

After this initial strain-induced formation of thin-layer sandwiches the number of thin-layer sandwiches is decreasing. For the samples TPU205 and TPU215 at the same time the number of sandwiches in the tail of the L-distribution is increasing. Because it is unlikely that at the same time one of the soft phases becomes more dense and the other becomes less dense, this observation shows that some of the thin-layer sandwiches are disrupted and, after that, the remnant hard domain may become member of a thick-layer sandwich. This transformation process is active up to a macroscopic strain e 1.6. Then the two materials have become so hard that they slip from the clamps of the tensile tester. 

For sample TPU 235 the transformation process is only dominant until s 0.75 (highest curve for r- 30 nm). After that even the number of thick-layer sandwiches starts to decrease, most probably because of destruction of their hard domains. As a 

---