General Non-Existence Theorem for Phase Transitions in One-Dimensional Systems with Short Range Interactions, and Physical Examples of Such Transitions
Download 370.08 Kb. Pdf ko'rish
|
1D-2
3.2. Chui–Weeks’s Model
We now proceed in order of increasing mathematical complexity and consider a model in which the transfer matrix has infinite size. The specific example we consider was proposed by Chui and Weeks (24) and is given by the following Hamiltonian: H N =J C N i=1 |h i − h i+1 | − K C N i=1 d h i , 0 . (15) 878 Cuesta and Sánchez This is a typical instance of the family of models called solid-on-solid (SOS) for surface growth, in which h i stands for the height above site i of the lattice; the reason for the name SOS is that overhangs are not allowed, i.e., the surface profile is single-valued. We will consider that heights can take on only integer values and that there is an impenetrable substrate, imposing h i \ 0. In this context, the first term represents the contribution of surface tension to the total energy, and the second one introduces an energy binding the surface to the substrate. As we will comment below, this is crucial for the model to exhibit a phase transition. Interestingly, these systems have often been considered as two-dimensional ones because of the fact that they represent interfacial phenomena on a plane, and therefore they have been considered not relevant for the 1D phase transition issue. We stress here that the fact that h i stands for a height does not change the 1D nature of the model, as it could equally well represent any other magnitude or internal degree of freedom, not associated to a physical dimension. Instead of following Chui and Weeks’s presentation, which is very simple but does not lead to explicit results, we resort to an alternative derivation proposed as Exercise 5.7 in Yeomans’s textbook. (25) We will not go here into the details of the derivation and quote only its main steps. A transfer matrix for the model is evidently ( T ) ij — e −bJ |i − j| [1+(e −bK − 1) d i, 0 ], i, j=1, 2,... . (16) Note that the matrix dimension is actually infinite, as announced, and stems from the fact that the amount of possible states ( heights) at any site of the lattice is infinite. It is also important to realize that in this case none of the entries in the matrix is zero, so we have a strictly positive matrix, although out of the scope of the theorems discussed above because of its infinite dimension. For simplicity, we introduce the notation w — e −bJ , o — e −bK . Then, by considering eigenvectors of the form v q — (k 0 , cos(q+h), cos(2q+h),...), (17) it is a matter of algebra to show that there is a continuous spectrum of eigenvalues, s( T )= 5 1 − w 1+w , 1+w 1 − w 6 . (18) Download 370.08 Kb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling