next up previous
Next: The AAA crystal Up: Phase diagram in Previous: Numerical Techniques

RESULTS

The results of the Gibbs-Duhem integration of the long rods are shown in table 5.11 and the coexistence densities are plotted in figure 5.10. The Gibbs ensemble results are displayed in table 5.12. As can be seen, the results of both methods agree quite well.

In figure 5.11 we have re-plotted the phase diagram in a way that makes comparison with the Onsager limit easier. In particular, we now use as our density variable

In these units the coexistence densities remain finite in the Onsager limit . In order to facilitate the extrapolation to the Onsager limit, we use rather than as the variable characterizing the aspect ratio of the rods. Figure 5.11 shows that the I-N coexistence curves extrapolate nicely to the Onsager limit at =3.29 and =4.19 [110], except for a few points on the nematic coexistence curve at high . Apparently, at these high lengths it is difficult to equilibrate the nematic phase completely. It is clear from figure 5.11 that already for rods with an aspect ratio less than =20 (or 0.05), the Onsager limit dominates the phase behavior. This result is not at all obvious as, for rods with this length, the Onsager assumption that all higher virial coefficients are negligible is not yet satisfied [111]. Yet, we cannot rule out the possibility that the discrepancy at higher (lower ) is an effect of higher virial coefficients. The agreement at lower values should then be attributed to a fortuitous cancellation of errors.

 

Figure: Equation of state for spherocylinders with aspect ratio =40.The dashed curves correspond to a compression whereas the solid curves denote an expansion. The different mechanically stable phases are indicated. All units as in figure 5.4.

Of course, the nematic to smectic and smectic to solid transition are more difficult to study, because they occur at higher density. This has two consequences: first of all, higher density means more particles in a box of a given (minimum) size, and secondly equilibration at high densities is slower. For =40 the equation-of-state is plotted in figure 5.12. We estimate that the reduced densities of the coexisting smectic and solid phases are =0.66 and =0.68 respectively. The nematic-smectic transition is estimated to take place at about =0.5. However, it is difficult to locate this transition accurately, as the smectic fluctuations in the nematic phase decay extremely slowly, even at lower densities. Poniewierski [112] has studied the nematic to smectic transition in the Onsager limit using bifurcation analysis. He found that the bifurcation point of the nematic-smectic transition is located at =0.531 which corresponds to a reduced density =0.46. This agrees with the decreasing trend in the transition density if one goes from =5 to =40. However, it also shows that the N-SmA transition for =40 is significantly different from the transition at .

In figure 5.2 we have combined our results for the phase behavior at large values with the low phase diagram discussed before. In order to give equal emphasis to all parts of the phase diagram, we have plotted the figure in the plane.

 

Figure:  Cartoon of the major difference between the AAA and the ABC stacking. Left: In the AAA stacking every particle is right on top of a particle in the layer below. Right: In the ABC stacking the layers are shifted.



next up previous
Next: The AAA crystal Up: Phase diagram in Previous: Numerical Techniques



Peter Bolhuis
Tue Sep 24 20:44:02 MDT 1996