Spin matrix representation in any arbitrary direction

PhysicsTruth
Messages
117
Reaction score
18
Homework Statement
Find the representation in which the component of spin along a direction ##\hat{n}## is diagonal, where ##\hat{n}## is a vector in the ##x-z## plane, making an angle ##\theta## with the ##z## axis.
Relevant Equations
##\sigma \cdot \hat{n} = \begin{pmatrix} cos(\theta) & sin(\theta)e^{-i\phi} \\ sin(\theta)e^{i\phi} & -cos(\theta) \end{pmatrix}##

##\sigma _z = \begin{pmatrix} 1 & 0\\ 0 & -1 \end{pmatrix}##
I've tried to use the 1st equation as a matrix to determine, but it clearly isn't a diagonal matrix. My guess is that I need to find the spin matrix along the direction ##\hat{n}##, but do I need to find the eigenstates of ##\sigma \cdot \hat{n}## first and check if they form a diagonal matrix or not? Can someone help me in figuring out how to proceed?
 
Physics news on Phys.org
You could find the eigenvectors of that matrix, but other than that I'm not sure what the question is asking.
 
  • Like
Likes vanhees71
The matrix you've found would then be diagonal in the basis of its eigenvectors. That might be what is intended.
 
  • Like
Likes vanhees71
PeroK said:
The matrix you've found would then be diagonal in the basis of its eigenvectors. That might be what is intended.
To be honest, the matrix doesn't have non-trivial solutions explicitly. It's only when one of the values is decided can the other value be determined. So, it might be diagonal only for a particular value of ##\theta## or ##\phi##. Also, ##\phi## hasn't been mentioned explicitly, so can't really proceed with that even.
 
PhysicsTruth said:
To be honest, the matrix doesn't have non-trivial solutions explicitly. It's only when one of the values is decided can the other value be determined. So, it might be diagonal only for a particular value of ##\theta## or ##\phi##. Also, ##\phi## hasn't been mentioned explicitly, so can't really proceed with that even.
It's manifestly Hermitian so must have an orthonormal eigenbasis. Moreover, it's physically identical to ##S_z## etc. So must have an eigenbasis.
 
  • Like
Likes PhysicsTruth and vanhees71
Another way is to think about how rotations act on spinors an tgen use the rotation which maps ##\vec{e}_z## to ##\vec{n}##.
 
  • Informative
Likes PhysicsTruth
vanhees71 said:
Another way is to think about how rotations act on spinors an tgen use the rotation which maps ##\vec{e}_z## to ##\vec{n}##.
Okay, so the rotation around any unit axis ##\hat{n}## of a spinor is given by-
##Icos(\theta /2) + i(\hat{n} \cdot \sigma)sin(\theta /2)##
So here, it's getting rotated about the ##y## axis, so should I get the representation as required if I use this rotation form?
 
Can someone just tell me how to write a Spin Matrix along any vector rotated by an angle ##\theta## from ##z## axis in terms of the eigenbasis of ##\sigma _z## Pauli spin matrix? Is it ##cos(\theta /2) |+> + sin(\theta /2) |->##, where ##|+>, |->## are eigen vectors of ##\sigma _z## ?
 
PhysicsTruth said:
Can someone just tell me how to write a Spin Matrix along any vector rotated by an angle ##\theta## from ##z## axis in terms of the eigenbasis of ##\sigma _z## Pauli spin matrix? Is it ##cos(\theta /2) |+> + sin(\theta /2) |->##, where ##|+>, |->## are eigen vectors of ##\sigma _z## ?
In that case, I can just perform an Unitary transformation to make it diagonal. So I need to know this.
 
  • #10
PhysicsTruth said:
Can someone just tell me how to write a Spin Matrix along any vector rotated by an angle ##\theta## from ##z## axis in terms of the eigenbasis of ##\sigma _z## Pauli spin matrix? Is it ##cos(\theta /2) |+> + sin(\theta /2) |->##, where ##|+>, |->## are eigen vectors of ##\sigma _z## ?
You already have the matrix in your original post. For the specific case of a direction in the x-z plane you have ##\phi =0##.
 
  • #11
PhysicsTruth said:
Can someone just tell me how to write a Spin Matrix along any vector rotated by an angle ##\theta## from ##z## axis in terms of the eigenbasis of ##\sigma _z## Pauli spin matrix? Is it ##cos(\theta /2) |+> + sin(\theta /2) |->##, where ##|+>, |->## are eigen vectors of ##\sigma _z## ?
Yes, that's the eigenvector corresponding to eigenvalue ##1##. You need the second eigenvector corresponding to eigenvalue ##-1##.
 
  • #12
PhysicsTruth said:
In that case, I can just perform an Unitary transformation to make it diagonal. So I need to know this.
You don't need to calculate the transformation to the eigenbasis, as you know it takes the matrix takes the same form as the Pauli matrix for ##S_z## in that basis.
 
  • #13
So, would the representation just be ##\begin{pmatrix} 1 & 0\\ 0 & -1 \end{pmatrix}## in that case?
 
  • #14
PhysicsTruth said:
So, would the representation just be ##\begin{pmatrix} 1 & 0\\ 0 & -1 \end{pmatrix}## in that case?
I did get this after a wholesome unitary transformation to the new eigenbasis, which is the same as the ##\sigma _z## matrix.
 
  • Like
Likes PeroK
  • #15
PhysicsTruth said:
So, would the representation just be ##\begin{pmatrix} 1 & 0\\ 0 & -1 \end{pmatrix}## in that case?
Yes, once you have the eigenvalues and vectors you know the matrix in that eigenbasis. Although, it can't hurt to grind out the transformation just to check!
 
  • Like
Likes PhysicsTruth
  • #16
PeroK said:
Yes, once you have the eigenvalues and vectors you know the matrix in that eigenbasis. Although, it can't hurt to grind out the transformation just to check!
Thanks. I have some problem to understand this intuitively. Can you just explain a bit as to why we end up getting this sole matrix when we change our basis and fix it along a direction?
 
  • #17
PhysicsTruth said:
Thanks. I have some problem to understand this intuitively. Can you just explain a bit as to why we end up getting this sole matrix when we change our basis and fix it along a direction?
The physical explanation is that the z-axis is an arbitrary choice. Any other direction is physically equivalent (and, indeed, could be chosen as the z-axis). The formalism of spin about any direction must be equivalent to the formalism about the z-axis.

Mathematically, the action of an operator is completely determined by its action on a basis. And, the action on its eigenbasis is "diagonal". You don't have to transform bases, you can study the action of the operator on that eigenbasis directly.
 
  • Like
Likes PhysicsTruth and vanhees71
Back
Top