All data points have been averaged nine different initial configurations. Error bars are included in the plots below.
Early-time structure: Elongated (but not fibrillar) polymer-rich domains (minor phase) in a LC-rich matrix (major phase).
Resulting intermediate/late-time structure: Noncircular polymer-rich domains (minor phase) in a LC-rich matrix (major phase). View phi and S profiles.
Evolution of degree average of PS and PO. Error bars included.
for PS (red): delta_phi(t) = phimax(t) - phimin(t)
for PO (blue): S(t) = Smax(t)
Both PS and PO well-established by t=5. View phi and S profiles.
#1: "old" k1 vs. "new" k1: using phi-based S(k).
"old" k1: R ~ t1/2 --> t1/3; "new" k1: R ~ t1/3 --> slower than t1/3.
We focus on our "new" k1 results. Initially, domain growth appears to have a t1/3 growth law (Model B). Growth slows down around t=20, but PS and PO both well-established by t=5. What causes the slowing down? Should we even get t1/3 behavior? Perhaps the initial behavior is transient and the only significant result is the eventual slowing down of domain growth due to the coupling of PS and PO. But why don't we see this "intermediate" behavior for quenches that are unstable wrt PS and PO, but more unstable wrt PO? Maybe the rate at which ordering occurs becomes faster than the rate at which PS occurs at t~20, causing domain growth to slow down. This would not be observed in quenches like E7-2 (unstable wrt PS and PO, but more unstable wrt PO), since for such a quench, the rate at which ordering happens already exceeds the rate at which PS happens. How can we check this?
Note: Initial growth behavior is actually slightly faster than t1/3. (See below.)
#2: "old" k1 vs. "new" k1: using S-based S(k).
Results similar to those of the phi-based results. See comments in #1 above.
#3: phi-based S(k) vs. S-based S(k): using "new" k1.
S-based data reveal similar growth behavior as seen from phi-based data for entire time range. Quench with similar resulting morphology, but under different quenching conditions (i.e., unstable to PS and PO, but more unstable to PO) does NOT show the same behavior. (See E7-2 runs. What about neighboring quenches? See F21-2: phi0=0.84 and F5-2: phi0=0.86.
(Remember that these power law fits are meant as guides and no way imply scaling behavior.)
#4: phi-based S(k) vs. S-based S(k): using "old" k1.
(Remember that these power law fits are meant as guides and no way imply scaling behavior.)
We can compare these results to those of a different lattice size:
To view directly the results of those of a different lattice size:
Jump to the individual results of the quench with (phi0, T, N) of:
Other links:
www.chem.ucla.edu/~aml/research.html
Last updated August 1, 1999.