Say we have two identical spin 3/2 particles in the same orbital state. What are the possible total spin? I know that there is a simple formula for adding angular momenta, but this breaks down when the particles occupy the same orbital state. Is there a formula to solve this problem for fermions of any spin?


For fermions, the total state has to be antisymmetric:

$$\Psi(x_1, x_2) = -\Psi(x_2, x_1) $$

where $\Psi$ can be factored into a spatial part and a spin part:

$$\Psi(x_1, x_2) = \psi(x_1, x_2)\chi_{1,2} $$

You've specified that the spatial part have the same orbital function, so:

$$ \psi_{\pm}(x_1, x_2) \equiv \frac 1 {\sqrt 2}[f(x_1)f(x_2) \pm f(x_2)f(x_1)]$$

where I've explicitly shown the (anti)symmetric form. Note that the antisymmetric part vanishes for all $x_i$ so the spatial part must be symmetric:

$$ \psi_+ = \frac 1 {\sqrt 2}[f(x_1)f(x_2) + f(x_2)f(x_1)]$$

That mean $\chi_{1, 2}$ must be an antisymmetric.

Since the particles are spin 3/2, we need to look at antisymmetric combinations$^1$ of $|m_1\rangle$ and $|m_2\rangle$ in terms of $|J, M\rangle$, the total spin state(s):

$$ |\frac 3 2\rangle|\frac 1 2\rangle - |\frac 1 2\rangle|\frac 3 2\rangle = \sqrt 2 |2, 2\rangle $$

$$ |\frac 3 2\rangle|-\frac 1 2\rangle - |-\frac 1 2\rangle|\frac 3 2\rangle = \sqrt 2 |2, 1\rangle $$

$$ |\frac 3 2\rangle|-\frac 3 2\rangle - |-\frac 3 2\rangle|\frac 3 2\rangle = |2,0\rangle + |0, 0\rangle $$

$$ |\frac 1 2\rangle|-\frac 1 2\rangle - |-\frac 1 2\rangle|\frac 1 2\rangle = |2,0\rangle - |0, 0\rangle $$

The remaining 2 nontrivial combinations can be found by subbing $m_1 \rightarrow -m1$, $m_2 \rightarrow -m_2$, and $M \rightarrow -M$.

From here you invert the equations to find the pure $|J, M\rangle$ states, and those are the eigenstates of total spin.

[1] To find the antisymmetric combinations of $m_1$ and $m_2$, you start with Clebsch-Gordan coefficients, $C_{\frac 3 2 m_1 \frac 3 2 m_2 J M}$. For example, the various:

$$c_{\frac 3 2 +\frac 3 2 \frac 3 2 -\frac 3 2 J 0} $$


$$ |\frac 3 2\rangle|-\frac 3 2\rangle= \frac 1 2 |0, 0\rangle + \sqrt{\frac 9{20}}|1, 0\rangle + \frac 1 2 |2, 0\rangle + \sqrt{\frac 1 {20}}|3, 0\rangle $$

Note that if you interchange $m_1$ and $m_2$:

$$ |-\frac 3 2\rangle|+\frac 3 2\rangle= -\frac 1 2 |0, 0\rangle + \sqrt{\frac 9{20}}|1, 0\rangle - \frac 1 2 |2, 0\rangle + \sqrt{\frac 1 {20}}|3, 0\rangle $$

So it is clear that the (anti)symmetric combinations have J odd (even).

The general principle is as follows: when taking the tensor product of 2 identical representations, you use Young diagrams. This gives you a totally symmetric combination, and a totally antisymmetric combination. The Hook length formula can then be used to compute the dimensions of these two irreducible representations. For spinors, vectors, and spin 3/2 (or 4-vectors), and spin-2, you get the following (respectively):

$$ {\bf 2} \otimes {\bf 2} = {\bf 3}_S + {\bf 1}_A $$ $$ {\bf 3} \otimes {\bf 3} = {\bf 5}_S + {\bf 3}_A $$ $$ {\bf 4} \otimes {\bf 4} = {\bf 10}_S + {\bf 6}_A $$ $$ {\bf 5} \otimes {\bf 5} = {\bf 15}_S + {\bf 20}_A $$

So in our spin 3/2 case where we are looking only at antisymmetric combinations, we have six dimensions corresponding to:

$$ {\bf 6} = {\bf 5} \oplus {\bf 1} $$

corresponding to spin-2 and spin-0, aka tensor and scalar (even $J$).

The symmetric combinations are:

$$ {\bf 10} = {\bf 7} \oplus {\bf 3} $$

corresponding to the spin-3 and spin-1 (vector) (odd $J$).

The thing about the Hook length formula is that it is all-powerful, and will give the dimension of any irreducible subspace over any field in any number of dimensions.

Perhaps it is too much? By inspection, if you combine 2 identical $n$ dimensional representations, the symmetric combination has $T_n$ dimensions, and the antisymmetric combination has $T_{n-1}$ dimensions, where $T_n$ are the Triangular Numbers:

$$ T_n \equiv \frac 1 2 n(n+1) $$

Moreover, for an odd (even) dimensional spin, the (anti)symmetric combinations are all even $J$, and for an even (odd) dimensional spin, the (anti)symmetric combinations are all odd $J$.

| cite | improve this answer | |
  • $\begingroup$ I have two questions. First, how did you write the total spin vectors? Was that with the tensor product between the |m1> |m2> vectors? $\endgroup$ – Michael O'Brien Dec 7 '18 at 17:18
  • $\begingroup$ Second, if we aren’t interested in the eigenstates in the |J,M> basis, is there a quick and dirty way to deduce the possibilities for total spin? $\endgroup$ – Michael O'Brien Dec 7 '18 at 17:20

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.