In Appendix A of this paper, the authors start from (equ. A1a and A1b)

$$\sigma \cdot \hat{n} |\hat{n},+\rangle = |\hat{n},+\rangle $$ $$\sigma \cdot \hat{n} |\hat{n},-\rangle = -|\hat{n},-\rangle $$

and claim that this system of linear equations is solved by (A2a and A2b):

$$|\hat{n},+\rangle = \cos(\frac{\theta}{2})e^{-i\phi/2}|+\rangle+\sin(\frac{\theta}{2})e^{i\phi/2}|-\rangle$$ $$|\hat{n},-\rangle = -\sin(\frac{\theta}{2})e^{-i\phi/2} |+\rangle + \cos(\frac{\theta}{2})e^{i\phi/2}|-\rangle$$

I do not get how they did this. In order to solve it by myself I assumed that what they meant by $\sigma \cdot \hat{n}$ was: $$\sigma \cdot \hat{n} = \begin{bmatrix} \cos(\phi) & \sin(\phi)e^{-i\theta} \\ \sin(\phi)e^{i\theta} & -\cos(\phi) \end{bmatrix}$$ as e.g. shown in the answer here. Therefore $|\hat{n},+\rangle$ and $|\hat{n},-\rangle$ should simply be the eigenvectors of this matrix. I computed the eigenvectors here, but this does not yield the same result as in the paper cited above (I tried trigonometric identities but this led me nowhere). As far as I understood $|+\rangle$ and $|-\rangle$ are supposed to be the eigenvectors of the Pauli matrix $\sigma_z$, since they are ''states of spin-up and -down along a specified direction, commonly $\hat{z}$''. [i.e. the vectors are supposed to be $|+\rangle = (1,0)$ and $|-\rangle = (0,1)$, which is the basis of the matrix I assumed above. Therefore I don't see where my error can be found.]

What did I do wrong? How can I derive $|\hat{n},+\rangle$ and $|\hat{n},-\rangle$ correctly?


1 Answer 1


First, note that your $\vec{\sigma}\cdot\hat{n}$ indicates that you are using $\phi$ for the polar angle and $\theta$ for the azimuthal one, while in the solution it is defined the opposite way.

Then, some hints:

  1. Using trigonometric identities you can get




  1. The eigenvectors must be normalized.

  2. You can multiply a vector by a gobal phase $e^{i\gamma}$.

I think you should be able to get the result with this, and don't forget the issue with the definition of angles!


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.