1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Correlation function, 1d polymer

  1. Apr 25, 2007 #1
    1. The problem statement, all variables and given/known data
    1D polymer, fixed segment length a
    If the angle between segment j and j+1 is 0, the energy is 0
    If the angle is pi the energy is +2J.

    Compute the correlation function [tex]<s_i s_{i+n}>[/tex], where [tex]s_j = \pm 1[/tex] denotes the direction of segment j

    Find the persistence length Lp, defined through
    [tex]<s_i s_{i+n}> = e^{-|n|a/Lp} [/tex]

    Find an expression for the end-to-end distance [tex]S(N) = <(x_N - x_0)^2>^{1/2}[/tex] as a function of temperature and the number of links N

    2. Relevant equations

    3. The attempt at a solution
    [tex]<s_i s_{i+n}> = \frac{ Tr s_i s_{i+n} e^{-\beta H} }{ Tr e^{-\beta H} }[/tex]

    But I don't know any hamiltonian? Or even what sort of trace to do.
    The problem sort of reminds me of the 'XY'-modell for spins on a 1d lattice, but I don't really understand how to make any use of that.
    Last edited: Apr 25, 2007
  2. jcsd
  3. Apr 26, 2007 #2


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    You can write down a Hamiltonian.
    It will be the sum of the energy of all the adjacent pairs, taking into account the rule they give. You want an expression that gives zero when [itex]s_j = s_{j+1}[/itex] and which gives 2J when [itex]s_j = -s_{j+1}[/itex]. This is simply [itex]J(1-s_j s_{j+1})[/itex].
  4. Apr 29, 2007 #3
    So, starting with a Hamiltonian like
    [tex]H = J \sum_{j=1}^{N-1} (1-s_j s_{j+1}) [/tex]
    and proceeding to calculate the partition function much the same way as for the 1d ising chain,
    [tex]Z = Tr e^{-\beta H} = \dots = 2 e^{-\beta J (N-1)} [2 cosh(\beta J)]^{N-1}[/tex]
    then noteing I can write
    [tex] <s_j s_{j+n}> = \frac{1}{Z \beta^n} \frac{d^n Z}{d J^n} [/tex]
    I end up with
    [tex] <s_j s_{j+n}> = tanh(\beta J)^n [/tex]
    is this really right? feels like I missed something

    thanks for the help!
    Last edited: Apr 29, 2007
  5. May 11, 2008 #4
    the last equation should be (tanh(\beta J))^n
  6. Oct 30, 2011 #5
    how can we calculate $$S^2(N) = <(x_N-x_0)^2>$$ in above case?
  7. Nov 1, 2011 #6
    How one can calculate

    S(N)2 = <(xN-x0)2>

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook