4
$\begingroup$

Have the correlation functions of the XY spin chain model,

\begin{equation} H=-\sum_l (J_x \sigma_l^x \sigma_{l+1}^x+J_x \sigma_l^y \sigma_{l+1}^y)-B\sum_l \sigma_l^z \end{equation}

been calculated using a functional partition function with source terms?

By functional partition function I mean the partition function plus source terms (or generating functional) which allow correlation functions such as $\langle \sigma_l^x \sigma_{l+1}^x \rangle$ to be calculated analytically:

\begin{equation} Z[\bar{\omega},\omega] = Tr \left[Exp \left[-\int_0^\beta d\tau H(a^{\dagger},a) - \int_0^\beta d\tau \sum_l (\bar{\omega} a + a^{\dagger} \omega) \right] \right] \end{equation}

where $H(a^{\dagger},a)$ is the XY Hamiltonian expressed in terms of spinless fermions (i.e. after a Jordan-Wigner transformation).

Edit: Anyone? An Ising chain would be fine too!

$\endgroup$
1
$\begingroup$

Some times it is not an efficient way to calculate the partition function using the functional integral. Even for a system of free fermions, we must have to evalute some complex Matsubara sums

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.