0

I know that if I have a single fermion then the expectation value of the spin along the $\hat{n}$ direction can be computed by $\sigma \cdot \hat{n}$, where $\sigma_i$ are the pauli matrices.

Now if I have a multiparticle system, is there a generalization of the pauli matrices (lets call them $\rho_i$) such that the expectation value of total spin along the $\hat{n}$ direction can be computed by $\rho \cdot \hat{n}$?

Qmechanic
  • 201,751
YankyL
  • 423

2 Answers2

0

Assuming you want an $N$ particle system of spin 1/2 fermions, then the corresponding operator is $\sum_{j=1}^N (\vec \sigma_j \cdot \hat n)$, where $\vec \sigma_j = \hat x \sigma_{jx}+ \hat y \sigma_{jy}+\hat z\sigma_{jz} $ are the Pauli operators for particle $j$. You need $2^N$ basis states, and the usual Pauli matrices will tell you what the corresponding matrix elements are. As an example $\langle \uparrow \downarrow | \sigma_{2x} |\uparrow \uparrow\rangle = 1$, Since the usual Pauli operator $\sigma_x$ flips the single spin, $\sigma_{2x}$ flips the second spin.

Typically, it is easier to use the operators to directly calculate the needed matrix elements rather than to write out $2^N \times 2^N$ matrices.

user200143
  • 2,470
0

There are two quantum numbers you will have to consider:

  1. intrinsic spin

  2. azimuthal

If you use the z axis, then the x and y axis' quantum numbers will be undefined as per the Pauli exclusion principle.

For two fermions there are four spin states:

αα, ββ, αβ, βα

The first two are symmetric, the last two are not symmetric or antisymmetric when electrons are exchanged so linear combination has to be made.

  1. spin paired total spin=0, antisymmetric

$\chi_{-}:=\frac{1}{\sqrt{2}}\left[\alpha(1)\beta(2)-\alpha(2)\beta(1)\right]$

  1. and three triplet states with total spin=1, symmetric

$\chi_{\alpha}:=\alpha(1)\alpha(2),\quad\chi_{\beta}:=\beta(1)\beta(2),\quad\chi_{+}:=\frac{1}{\sqrt{2}}\left[\alpha(1)\beta(2)+\alpha(2)\beta(1)\right]$

These all have the same energy level.