I'm asked to prove this relation: $$\left [\textbf{A},\textbf{L} \cdot \textbf{B} \right]=-i \hbar \textbf{A} \times \textbf{B} + L_i \left [\textbf{A},B_i \right].$$ Where $\textbf{A}$ and $\textbf{B}$ are vector operators that satisfy $$\left[L_i,A_j \right]=i \hbar A_k\varepsilon_{ijk},$$ and $\textbf{L}$ is the angular momentum operator.

This is easy using $\left[A_k,L_j B_j \right]=L_j \left[A_k, B_j \right]+ \left[A_k, L_j \right]B_j=L_j \left[A_k, B_j \right]-i \hbar A_l\varepsilon_{jkl}Bj$. Then I'm asked to use this relation to show that $$\textbf{r}(\textbf{L} \cdot\textbf{p})-\textbf{p}(\textbf{L} \cdot\textbf{r})=0.$$ Here is where I'm stuck. Setting $\textbf{A}=\textbf{r}$ and $\textbf{B}=\textbf{p}$ I am able to show that $\textbf{r}(\textbf{L} \cdot\textbf{p})-(\textbf{L} \cdot\textbf{p})\textbf{r}=0.$ With $\textbf{B}=\textbf{r}$ and $\textbf{A}=\textbf{p}$ I get $\textbf{p}(\textbf{L} \cdot\textbf{r})-(\textbf{L} \cdot\textbf{r})\textbf{p}=0.$ But this gets me nowhere.

Am I missing something? The formula is similar to an identity involving $\textbf{L} \times (\textbf{r} \times \textbf{p})$. But I think that the result has to be more or less immediate as it is stated as a corollary to the first relation.

  • $\begingroup$ If you subtract the two terms you get $Req + (L.r)p - (L.p)r = 0$. Where Req is the quantity you have to show to be equal to zero. The remaining part can be expressed as $L \times (r \times p)$ which is zero, since $L = r \times p$. I think this process is fine. $\endgroup$
    – M111
    Nov 26, 2017 at 21:19
  • $\begingroup$ $\textbf{L} \times \textbf{L}=i \hbar \textbf{L}$ is not zero. And Req is not equal to that, it is missing a commutator. $ a × (b × c) = b (a \cdot c) − (a \cdot b) c + [ a_j , b ] c_j$ $\endgroup$ Nov 26, 2017 at 21:33
  • $\begingroup$ Still, I have the feeling that using that identity is too much work. I would have to prove it, together with the fact that $\textbf{L} \times \textbf{L}=i \hbar \textbf{L}$. This seems more work than simply proving $\textbf{r}(\textbf{L} \cdot\textbf{p})-\textbf{p}(\textbf{L} \cdot\textbf{r})=0$ by hand. I was hoping there was some kind of neat trick here. $\endgroup$ Nov 26, 2017 at 21:39

1 Answer 1


If you're allowed to assume that $\textbf{L}=\textbf{r} \times \textbf{p}$ then the equation you're trying to prove is in some sense totally trivial, because both terms vanish identically: $$\textbf{L} \cdot \textbf{r} = \textbf{L} \cdot \textbf{p}=0$$ This is very easy to prove by noting that the usual vector equality $\textbf{L}=\textbf{r} \times \textbf{p}=-\textbf{p} \times \textbf{r}$ still holds for those operators.

However if you're not allowed to assume that, it is not true in general that the dot product of $\textbf{L}$ and an arbitrary vector operator vanishes (the obvious counterxample is $\textbf{L}\cdot \textbf{L}\neq 0)$. Therefore, to avoid being blinded by preconceived notions, one should try to prove the following: $$\textbf{A}(\textbf{L} \cdot\textbf{B})-\textbf{B}(\textbf{L} \cdot\textbf{A})=0$$ for any two vector operators $\textbf{A}$ and $\textbf{B}$. However this is not true in general, the counterexample given by taking $\textbf{A}=\textbf{r}$ and $\textbf{B}=\textbf{L}$, the usual $\textbf{L}$. Then $$\textbf{A}(\textbf{L} \cdot\textbf{B})-\textbf{B}(\textbf{L} \cdot\textbf{A})=\textbf{r}\textbf{L}^2\neq 0$$ Therefore I suspect that the first part of the question is indeed what they were looking for. The result is simple enough to be immediate even without resorting to any lemmas.


Your Answer

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

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