2
$\begingroup$

$$c\left(\alpha _i\right.{\cdot P + \beta mc) \psi = E \psi } $$ From the above dirac equation it can be shown for zero momenta that spin and antimatter are associated with $\beta $.

On the other hand, how are each of the momenta $P_x$,$P_y$,$P_z$ associated with $\alpha _1$,$\alpha _2$,$\alpha _3$? I mean does each $\alpha _1$,$\alpha _2$,$\alpha _3$ correspond directly with $P_x$,$P_y$,$P_z$ ?

OR

are they related to the Four-Momenta $P_4$ of Minkowski space, and can I write $P_4$ as

$$ \left( \begin{array}{c} E/c \\ p_x \\ p_y \\ p_z \end{array} \right) $$

I understand they are covariant, which follows naturally from the Dirac equation, but I just can't see how $\alpha _1$,$\alpha _2$,$\alpha _3$ operate to yield the momentum $P_x$,$P_y$,$P_z$.

$\endgroup$

2 Answers 2

2
$\begingroup$

Let's consider a massless Dirac fermion. In this case, we get rid of $\beta$, only three $\alpha_i$ to be determined. These matrices should anti-commute with one another and square to $\bf 1$. Pauli matrices will suffice and they are genuine spins! Then you have the correspondence: $$p^\mu \rightarrow \sigma^\mu$$ where $\sigma^0=\bf 1$.

Now consider a finite mass. This time, three 2x2 Pauli matrices will not work since we need four $(\alpha;\beta)$. The solution is enlarging the matrix dimension to 4x4 by taking tensor product $\Gamma=\sigma\otimes\tau$. We have to introduce a pseudo-spin $\tau$, which corresponds to particle-hole index. You can still make some sort of indentification of spin and momentum by looking at the $\sigma$ part.

$\endgroup$
1
$\begingroup$

$\alpha_1$,$\alpha_2$,$\alpha_3$ equal operators of components of velocity up to a constant factor (see Dirac, Principles of quantum mechanics, or just calculate commutators of coordinate operators with the Hamiltonian).

$\endgroup$

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.