In a book of "LES HOUCHES - Critical Phenomena, Random systems, Gauge theories" the author Frolich says that:


In two dimensions, the mean energy of an isolated point defect in a square area of diameter $L$ is proportional to $\log(L)$. The total number of possible positions is proportional to $L^2$, i.e. the entropy grows logarithmically in $L$. Hence the free energy behaves like:

$F=E-TS \sim \textrm{const.} \log(L)-kT \textrm{const.}^\prime \log(L)$

Thus, for $T$ large enough, a dilute system of bound point defects becomes unstable in the thermodynamic limit, i.e. defects unbind and form a plasma


In three dimensions, dislocations are line defects with a self-energy roughly proportional to their lenght, $L$. In a cubic area of diameter $\sim \textrm{const.} L$ the number of possible configurations of a single dislocation loop of lenght $L$ is clearly proportional to $\exp(\textrm{const.} L)$, so the entropy grows linearly in $L$. The free energy thus behaves like

$F \sim \textrm{const.} L - kT \textrm{const.}^\prime L$

I have not understood the following points:

  1. What is the qualitative argument that says that the mean energy of a point defect in a square is $∼\log(L)$?

  2. The same question of 1. for the 3D case.

  3. How he derives the proportionality of $\exp(\textrm{const.} L)$ ?

  4. The same question of 3. for the 2D case

  5. How can he says that for $T$ large the system of point defects is unstable in the thermodynamic limit?

  6. In 2D he speaks of "mean energy" in 3D of "self-energy". Why does he make this difference?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.