In the second quantization formalism of quantum chemistry, according to the book 1 on page 43, the creation operators $a_{p,\beta}^{\dagger},a_{p,\beta}^{\dagger}$ satisfy the relations

  • $[\hat S_\pm,a_{pm_s}^{\dagger}]=\sqrt{\frac{3}{4}-m_s(m_s\pm1)}a_{p,m_s\pm1}^{\dagger}$ (1)

  • $[\hat S_z,a_{pm_s}^{\dagger}]=m_s a_{pm_s}^{\dagger}$ (2).

The following form of the spin functions was used according to the author:

  • $\hat S_+ = \sum_p a^\dagger_{p\alpha}a_{p\beta}$ (3)

  • $\hat S_- = \sum_p a^\dagger_{p\beta}a_{p\alpha}$ (4)

  • $\hat S_z = \frac{1}{2}\sum_p (a^\dagger_{p\alpha}a_{p\alpha}-a^\dagger_{p\beta}a_{p\beta})$ (5)

By comparing with the equations which define spin tensor operators:

  • $[\hat S_\pm,\hat T^{S,M}]=\sqrt{S(S+1)-M(M\pm1)}\hat T^{S,M\pm1}$ (6)

  • $[\hat S_z,\hat T^{S,M}]=M\hat T^{S,M}$ (7)

it follows that the creation operators above are spin tensor operators.

How do equations (1) and (2) follow from the form of the spin operators given in equation (3)-(5)?

  • 1
    $\begingroup$ Have you tried to just compute those commutators directly using the definitions by plugging in for the S operators? $\endgroup$
    – march
    Dec 15, 2021 at 18:30
  • $\begingroup$ I have tried it, but was stuck. So thanks for posting the solution :-) $\endgroup$
    – Jadzia
    Dec 16, 2021 at 11:21

1 Answer 1


You can just compute the commutators directly. For instance, \begin{align} [\hat S_z,\hat{a}_{qm}^{\dagger}] &= \left[ \frac{1}{2}\sum_p (\hat{a}^\dagger_{p\alpha}\hat{a}_{p\alpha}-\hat{a}^\dagger_{p\beta}\hat{a}_{p\beta}) ,\hat{a}_{qm}^{\dagger} \right] \\&= \frac{1}{2}\sum_p \left( \left[\hat{a}^\dagger_{p\alpha}\hat{a}_{p\alpha},\hat{a}_{qm}^{\dagger}\right] - \left[\hat{a}^\dagger_{p\beta}\hat{a}_{p\beta},\hat{a}_{qm}^{\dagger}\right] \right) \\&= \frac{1}{2}\sum_p \left( \hat{a}^\dagger_{p\alpha}\left[\hat{a}_{p\alpha},\hat{a}_{qm}^{\dagger}\right] + \left[\hat{a}^\dagger_{p\alpha},\hat{a}_{qm}^{\dagger}\right]\hat{a}_{p\alpha} - \left( \hat{a}^\dagger_{p\beta}\left[\hat{a}_{p\beta},\hat{a}_{qm}^{\dagger}\right] +\left[\hat{a}^\dagger_{p\beta},\hat{a}_{qm}^{\dagger}\right]\hat{a}_{p\beta} \right) \right) \\&= \frac{1}{2}\sum_p \left( \hat{a}^\dagger_{p\alpha}\delta_{qp}\delta_{\alpha m} + 0 - \left( \hat{a}^\dagger_{p\beta}\delta_{qp}\delta_{\beta m} +0 \right) \right) \\&= \frac{1}{2}\sum_p \left( \hat{a}^\dagger_{p\alpha}\delta_{qp}\delta_{\alpha m} - \hat{a}^\dagger_{p\beta}\delta_{qp}\delta_{\beta m} \right) \end{align} Then, performing the sum over $p$, this becomes \begin{align} [\hat S_z,\hat{a}_{qm}^{\dagger}] &= \frac{1}{2} \left( \hat{a}^\dagger_{q\alpha}\delta_{\alpha m} - \hat{a}^\dagger_{q\beta}\delta_{\beta m} \right)\,. \end{align} Note that if $m=\alpha$, then we get \begin{align} [\hat S_z,\hat{a}_{qm}^{\dagger}] &= \frac{1}{2} \hat{a}^\dagger_{q,\alpha}\,, \end{align} and if $m=\beta$, then we get \begin{align} [\hat S_z,\hat{a}_{qm}^{\dagger}] &= -\frac{1}{2} \hat{a}^\dagger_{q,\beta}\,, \end{align} indicating that $\alpha$ is acting as an $m=1/2$ quantum number and that $\beta$ is acting as a $m=-1/2$ quantum number, so we may as well relabel things so that $\alpha=1/2$ and $\beta = -1/2$. The other commutator is similar, although the calculation is more involved.

Note that, in the calculation above, we have used some standard commutator identities, which can be proved straight-forwardly by expanding out the commutators. These are \begin{align} [\hat{a},\hat{a}]&=0\,, \\ [\hat{a}+\hat{b},\hat{c}]&=[\hat{a},\hat{c}]+[\hat{b},\hat{c}]\,, \\ [\hat{a}\hat{b},\hat{c}]&=\hat{a}[\hat{b},\hat{c}]+[\hat{a},\hat{c}]\hat{b}\,. \end{align} In addition, we have also used the facts about the raising and lowering operators, where \begin{align*} \left[\hat{a}_{p\alpha},\hat{a}_{qm}^{\dagger}\right] = \delta_{pq}\delta_{\alpha m}\,. \end{align*}

  • $\begingroup$ Thank you, I can see now that the second equation of my initial post holds. I still cannot see the first equation (1) from my post. Somehow I get a different result. $\endgroup$
    – Jadzia
    Dec 16, 2021 at 13:33
  • $\begingroup$ Would you be able to show the first equation (1) as well? $\endgroup$
    – Jadzia
    Dec 17, 2021 at 11:36

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.