0
$\begingroup$

I understand that the Quantum Heisenberg XXZ model in 1D has the form:

$$\hat H = \frac{1}{2} \sum_{j=1}^{N} (J_x \sigma_j^x \sigma_{j+1}^x + J_y \sigma_j^y \sigma_{j+1}^y + J_z \sigma_j^z \sigma_{j+1}^z + h\sigma_j^{z})$$

with the term $J_z \sigma_j^z \sigma_{j+1}^z$ disfavoring/favoring nearby aligned spins depends on the sign of $\hat H$. However, I was wondering if there's a name for a model with the $J_z \sigma_j^z \sigma_{j+1}^z$ replaced by something that only cares for up-up nearby spins, but don't care about down-down alignment? Something like that could be implemented by a $J_z n_j n_{j+1}$ for example, with $n_j$ as the number operator at site $j$.

If there's a solution for it that'd be even better. Thank you!

$\endgroup$
2
  • 2
    $\begingroup$ By adjusting $h$ accordingly, you should recover your model, assuming what you call $n=\frac{\sigma_z+1}2$ $\endgroup$
    – LPZ
    Feb 17 at 23:13
  • $\begingroup$ I think it's $n = \sigma_z + \frac{1}{2}$, but in general you're correct. Thank you! $\endgroup$
    – Kim Dong
    Feb 18 at 13:59

1 Answer 1

1
$\begingroup$

As pointed out by @LPZ, the term you mention is \begin{equation} J_{\mathrm{new}}(4\sigma^z_i\sigma^z_{j}+2\sigma^z_i+2\sigma^z_j+1)/4, \end{equation} and therefore, if you plug this in to where you have $\sigma^z_i\sigma^z_j$ now and calculate the result, it ends up being just another XXZ model with $J_z^{\prime}=J_{\mathrm{new}}$ and $h^{\prime}=h+J_{\mathrm{new}}$, with a constant $NJ_{\mathrm{new}}/8$ shift in the baseline energy. So, that slightly different version ultimately is just the same as the usual XXZ Hamiltonian.

XXZ chains are known to have Bethe Ansatz solutions (if $h=0$), which is a beast on its own. There are many papers on it so I'd recommend searching for them.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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