5
$\begingroup$

We are supposed to give a matrix representation of $L\cdot S$ for an electron with $l=1$ and $s=\frac{1}{2}$.

I read $L\cdot S$ as $L \otimes S$. Is this correct? Then we would have e.g. for

$L\otimes S (|1,1\rangle \otimes |1/2,1/2\rangle) = L |1,1\rangle \otimes S|1/2,1/2\rangle $ $= \sqrt{2} \hbar |1,1\rangle \otimes \sqrt{\frac{3}{4}} \hbar |1/2,1/2 \rangle = \sqrt{\frac{3}{2}}\hbar^2 |1,1\rangle \otimes |1/2,1/2\rangle $.

Is this correction correct? In that case should I proceed in this way with all the other basis vectors and write the eigenvalues down the diagonal in a matrix?

$\endgroup$
  • 3
    $\begingroup$ ${\bf L}$ and ${\bf S}$ are vectors. ${\bf L} \cdot {\bf S}$ is a scalar. What is $L\otimes S$? If you want to use the tensor product of operators, the correct expression is ${\bf L} \cdot {\bf S} = \sum_{k=1}^3 L_i \otimes S_i$. $\endgroup$ – Valter Moretti Jan 17 '14 at 22:03
  • $\begingroup$ so you are saying that this is nonsense? well, do you know how to get a matrix representation for $L \cdot S$? $\endgroup$ – Xin Wang Jan 17 '14 at 22:07
  • $\begingroup$ Yes: replace each operator $L_i$ and $S_i$ for the corresponding matrix. Use the fact that $l=1$ and $s=1/2$ to find these matrices in textbooks. $\endgroup$ – Valter Moretti Jan 17 '14 at 22:08
  • $\begingroup$ I don't understand this. If we say that $J=L+S$, does this not mean that $J=L \otimes Id + Id \otimes S$? I think I have not understood the whole concept of addition of angular momenta. What is the correct mathematical notation for $J=L+S$? $\endgroup$ – Xin Wang Jan 17 '14 at 22:11
  • 2
    $\begingroup$ ${\bf J}= {\bf L} + {\bf S}$ actually means: $J_k = L_k \otimes I_{spin} + I_{orb}\otimes S_k$ for $k=1,2,3$. $\endgroup$ – Valter Moretti Jan 17 '14 at 22:14
8
$\begingroup$

There are two problems to deal with which must be disentangled to solve problems like these.

  • Both angular momentum operators are vector operators, so in some sense they "take values" in $\mathbb R^3$; you are being asked for their dot product, which should be taken within that copy of $\mathbb R^3$. You would have the same problem if you were asked to calculate the dot product $\mathbf r\cdot\mathbf p$ for a single particle without spin.

  • The orbital and spin angular momentum operators act on the two different factors of a tensor product of Hilbet spaces. Thus any (operator) product of a scalar orbital operator with a scalar spin operator should be interpreted as a tensor product. You would have the same problem if you were asked to calculate the product $L^2S^2$, which would need to be interpreted as $L^2\otimes S^2$.

Thus, in your case, you must read $L\cdot S$ as $$ \mathbf{L}\cdot \mathbf{S}=\sum_{i=1}^3L_iS_i=\sum_{i=1}^3L_i\otimes S_i. $$ To compute the matrix representation of this, you should begin with the matrix representation of each $L_i$ and $S_i$. You then compute the tensor product matrices $L_i\otimes S_i$. Finally, you add all of those matrices together to get the final result.

This is all much clearer with an example. The $z$ component, for example, is easy, since each matrix is given by $$ L_z=\hbar\begin{pmatrix}1&0&0\\0&0&0\\0&0&-1\end{pmatrix} \quad\text{and}\quad S_z=\frac\hbar 2 \begin{pmatrix}1&0\\0&-1\end{pmatrix}, $$ in the bases $\{|1\rangle,|0\rangle,|-1\rangle\}$ and $\{|\tfrac12\rangle,|-\tfrac12\rangle\}$ respectively. The tensor product matrix, then, in the basis $\{|1\rangle\otimes|\tfrac12\rangle ,|0\rangle\otimes|\tfrac12\rangle ,|-1\rangle\otimes|\tfrac12\rangle , |1\rangle\otimes|-\tfrac12\rangle ,|0\rangle\otimes|-\tfrac12\rangle ,|-1\rangle\otimes|-\tfrac12\rangle \}$, is given by $$ L_z\otimes S_z=\frac{\hbar^2} 2 \begin{pmatrix} 1\begin{pmatrix}1&0&0\\0&0&0\\0&0&-1\end{pmatrix} & 0\begin{pmatrix}1&0&0\\0&0&0\\0&0&-1\end{pmatrix} \\ 0\begin{pmatrix}1&0&0\\0&0&0\\0&0&-1\end{pmatrix} & -1\begin{pmatrix}1&0&0\\0&0&0\\0&0&-1\end{pmatrix} \end{pmatrix} =\frac{\hbar^2} 2 \begin{pmatrix} 1&0&0& 0&0&0\\0&0&0&0&0&0\\0&0&-1&0&0&0\\ 0&0&0&-1&0&0\\0&0&0&0&0&0\\0&0& 0&0&0&1 \end{pmatrix}. $$ This procedure should be repeated with both the $x$ and the $y$ components. Each of those will yield a six-by-six matrix (in this case). To get your final answer you should add all three matrices.

$\endgroup$

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.