there are a few answered questions regarding the commutator of any two 3D angular momentum operator components $[L_i, L_j]$ , however, I am trying to go through fully using index notation so that I can arrive at the generic expression $$L_k=i \hbar \epsilon_{kij}x_ip_j$$ where $x$ and $p$ are the position and momentum operators in 3D.

Using $$[AB,C] = A[B,C] + [A,C]B$$ and cancelling out self-commutators, as well as using the canonical commutation relations I come to this line (skipped some steps):

$$[L_i, L_j] = \epsilon_{iab}\epsilon_{jcd}[x_ap_b,x_cp_d] = ... = \epsilon_{iab}\epsilon_{jcd}(x_ap_d\delta _{bc} - x_bp_c\delta _{ad}) $$

Now if I try to contract the levi-civitas for each of the two terms in the brackets, I just get more kronecker deltas, with which I don't know what to do. If someone is feeling really generous today, could you go through the final steps to the solution so that I can understand what is happening (all the answers I have found just skip this assuming the reader will understand). Thank you!

  • $\begingroup$ I believe your question is answered here: physics.stackexchange.com/questions/65947/… $\endgroup$
    – user83548
    Oct 30, 2015 at 18:09
  • $\begingroup$ It kind of is, except for the part where the person answering says "I suggest you work out the missing parts to understand how this levi-civita business works." Which is where I am stuck $\endgroup$ Oct 30, 2015 at 18:20

1 Answer 1


You have mixed indices in the end of your line. Correctly: $$-\epsilon_{iab}\epsilon_{jcd}(x_ap_d\delta _{bc} - x_cp_b\delta _{ad}).$$ So further, $-\epsilon_{iab}\epsilon_{jcd}(x_ap_d\delta _{bc} - x_cp_b\delta _{ad})=\epsilon_{iab}\epsilon_{jcd}x_cp_b\delta _{ad}-\epsilon_{iab}\epsilon_{jcd}x_ap_d\delta _{bc}=\\=\epsilon_{idb}\epsilon_{jcd}x_cp_b - \epsilon_{iac}\epsilon_{jcd}x_ap_d=\epsilon_{iac}\epsilon_{jdc}x_ap_d-\epsilon_{ibd}\epsilon_{jcd}x_cp_b=\\=(\delta _{ij}\delta _{ad}-\delta _{id}\delta _{aj})x_ap_d-(\delta _{ij}\delta _{bc}-\delta _{ic}\delta _{bj})x_cp_b=\\=\delta _{ij}x_ap_a-x_jp_i-\delta _{ij}x_bp_b+x_ip_j=x_ip_j-x_jp_i=\epsilon_{ijk}L_k.$

  • $\begingroup$ Thank you so much! One final question - why does the epsilon appear at the end? I understand it's role but I don't understand the jump. Edit: so there is no epsilon there? $\endgroup$ Oct 30, 2015 at 18:46
  • $\begingroup$ It is obtained by comparing the expression vector product and our record in index form. For example, for a specific index values $(i=1, j=2)$ we obtain: $x_1p_2-x_2p_1=\epsilon_{123}x_1p_2+\epsilon_{213}x_2p_1=[\mathbf{x} \times \mathbf{p}]_3=L_3$ $\endgroup$
    – kw_artem
    Oct 30, 2015 at 19:15
  • $\begingroup$ It can be seen that the first two of the levi-civitas index are permuted in the same manner as the indices of the $x$ and $p$ operators. $\endgroup$
    – kw_artem
    Oct 30, 2015 at 19:22

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.