SO I have a hamiltonian: $$H=\alpha J_1\cdot J_2$$ And I am asked to find the eigenstates and eigenvalues of this Hamiltonian in terms of products of the eigenstates of the z components of the individual spins.

The wording of this question is kind of confusing me. I could solve it by rewriting it in terms of $(J_1+J_2)^2,J_1^2,J_2^2$ and then find the eigenvalues, but that is not in terms of the z components. How should I go about doing this? Thanks


The question is essentially just asking you to perform Clebsch-Gordan (CG) decomposition, namely a change of basis on the Hilbert space $\mathcal H_{j_1}\otimes\mathcal H_{j_2}$ of the composite system of spins.

Recall that for each of the Hilbert spaces $\mathcal H_{j_1}$ and $\mathcal H_{j_1}$ there exist orthonormal bases of eigenvectors of $\mathbf J_1^2, J_1^z$ and $\mathbf J_2^2, J_2^z$ respectively, and these bases are \begin{align} \text{for $\mathcal H_{j_1}$}&:\qquad \{|j_1, m_1\rangle\,|\,m_1=-j_1,-j_1+1,\dots,j_1-1, j_1\} \\ \text{for $\mathcal H_{j_2}$}&:\qquad \{|j_2, m_2\rangle\,|\,m_2=-j_2,-j_2+1,\dots,j_2-1, j_2\} \end{align} It follows that an orthonormal basis of eigenvectors for the composite Hilbert space $\mathcal H_{j_1}\otimes\mathcal H_{j_2}$ consists of the set of all tensor products of these basis elements; \begin{align} \text{for $\mathcal H_{j_1}\otimes\mathcal H_{j_2}$}:\qquad \{|j_1, m_1\rangle|j_2,m_2\rangle\,|\,-j_1\leq m_1\leq j_1, -j_2\leq m_2\leq j_2\} \tag{$\star$} \end{align} This is what being referred to as

"...products of eigenstates of the z components of individual spins."

So once you've found the eigenstates of the Hamiltonian, which you will presumably write in terms of the basis \begin{align} \{|j,m,j_1,j_2\rangle\}, \end{align} you just need to decompose them into the tensor product basis in $(\star)$ above.

  • $\begingroup$ So how could I do this generally, given J. I do not know what spin they have. The coupled would go from $j_1+j_2,...,|j_1-j_2|$. $\endgroup$ – yankeefan11 Mar 30 '14 at 1:15
  • $\begingroup$ @yankeefan11 To do this for a given $j_1$ and $j_2$, you either look in a CG table, or you compute the CG coefficients yourself. $\endgroup$ – joshphysics Mar 30 '14 at 1:16
  • $\begingroup$ Yeah, but I am not given $j_1,j_2$ $\endgroup$ – yankeefan11 Mar 30 '14 at 1:17
  • $\begingroup$ @yankeefan11 As far as I am aware, no closed form expression for CG coefficients for general $j_1$ and $j_2$ exists, hence the need for a table. $\endgroup$ – joshphysics Mar 30 '14 at 1:22
  • $\begingroup$ Thats why I am confused on how to answer this question. I could solve it given $J_1,J_2$, but I'm not given that $\endgroup$ – yankeefan11 Mar 30 '14 at 1:25

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.