0
$\begingroup$

I am creating an operator group from representation of spin 1 operators $$J_{x} = \frac{1}{\sqrt{2}}\left(\begin{array}{ccc} 0 & 1 & 0\\ 1 & 0 & 1\\ 0 & 1 & 0 \end{array} \right)\ \text{, } J_{y} = \frac{i}{\sqrt{2}}\left(\begin{array}{ccc} 0 & -1 & 0\\ 1 & 0 & -1\\ 0 & 1 & 0 \end{array} \right)\ \text{, } J_{z} = \left(\begin{array}{ccc} 1 & 0 & 0\\ 0 & 0 & 0\\ 0 & 0 & -1 \end{array} \right)$$ $$\hat{J}_{\mu} = \sum\limits_{m=-1,0,1}(J_{\mu})_{mn}\hat{a}^{\dagger}_{m}\hat{a}_{n},$$ and I want to find eigenstate of $\hat{J}_{z}$ (or other two) in Fock space with fixed number of particles $N$. For example, I have found that the eigenstate of $J_{z}$ should have the form $$\left| J_{z}\right> = \sum\limits_{k=1/2(N-M)}^{N-M}C_{k}\left|N-k, M-N+2k,N-M-k \right>$$ with eigenstate $M$ and coefficients $C_{k}$. Due to $SU(2)$ structure of the spin operators I suppose that $C_{k}$ are not random numbers, but there is some theory behind them. Is there a generall way to determine those numbers (something similar to Clebsch-Gordan coefficients?) or to find the eigenstates?

$\endgroup$
  • $\begingroup$ I am not sure what the three numbers in your ket state represent. How could a spin-1 N-particle state have only 3 components? In the mean time, please see my proposed answer and see if I have understood you correctly. $\endgroup$ – Pooya Nov 20 '14 at 4:01
  • $\begingroup$ Also see a similar question. $\endgroup$ – Pooya Nov 20 '14 at 18:57
1
$\begingroup$

To find the spin eigenstates corresponding to a multi-particle state, all one needs to do is build the appropriate multi-particle spin operator using the direct product, e.g. $$J_z^{(2)}=J_z \otimes 1 + 1\otimes J_z \\ J_z^{(3)}=J_z \otimes 1 \otimes 1 + 1\otimes J_z \otimes 1 + 1\otimes 1\otimes J_z $$ and then diagonalize the resulting operator to find the eigenstates. For the case of $n=2$ one has $$ J_z^{(2)}=diag(2,1,0,1,0,-1,0,-1,-2) $$ which happens to be already in diagonal form, therefore the eigenstates are trivial: $\langle j \vert \psi_i \rangle = \delta_{ij}$. In fact, because $J_z$ is diagonal, $J_z^{(n)}$ will also be diagonal for any $n$, therefore the corresponding eigenstates would be $\langle j \vert \psi_i \rangle = \delta_{ij}$, where $1 \le i,j \le 3^n$.

$\endgroup$

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.