1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Commuting operators

  1. Jun 27, 2005 #1
    I was wondering: is every eigenstate of L^2 also an eigenstate of Lz?

    I know that commuting operators have the same eigenfunctions but if [A,B] = 0 and a is a degenerate eigenfunction of A the the corresponding eigenfunctions of A are not always eigenfunctions of B.
     
  2. jcsd
  3. Jun 27, 2005 #2

    Galileo

    User Avatar
    Science Advisor
    Homework Helper

    Not necessarily. What [A,B]=0 means (for hermitian operators) is that there EXISTS a basis of eigenvectors common to A and B.
    If A|a>=a|a> and a is a nondegenerate eigenvalue of A, then |a> is also an eigenvector of B (this is easy to prove). If a is degenerate, then you can find an orthonormal basis in the eigenspace of a consisting of eigenvectors common to A and B.
    So not every eigenstate of L^2 is an eigenstate of Lz.
     
  4. Jun 27, 2005 #3

    dextercioby

    User Avatar
    Science Advisor
    Homework Helper

    You can make it a nice discussion in the following way.

    [tex] \hat{J}^{2} ,\hat{J}_{3} [/tex]

    form a complete system of commuting observables.It's not difficult to show that.
    Therefore,a spectral pair [itex]\left(\hbar^{2} j(j+1), \hbar m_{j}\right) [/itex] determine,up to a phase factor,a vector from the irreducible space of the irreducible finite dimensional linear representation of the angular momentum algebra* (which is isomorphic to [itex] su(2) [/itex] which is isomorphic to [itex] so(3) [/itex]).This is an eigenvector to both operators.

    *One can prove that an eigensubspace of [itex] \hat{J}^{2} [/itex] associated to the nondegenerate spectral value [itex] \hbar^{2} j(j+1) [/itex] is invariant to the action of all the linops of the representation of the ang.mom.algebra and does not admit any nontrivial invariant subspaces,therefore it is a subspace of an irred.representation of the ang.mom.algebra.

    So all eigenstates of [itex] \hat{J}^{2} [/itex] are eigenstates of [itex] \hat{J}_{3} [/itex] in the simple case of uniparticle systems.

    Daniel.

    P.S.For the particular case of orbital ang.mom.ops,the things are a bit easier,as one has the particular realization [itex] \mathcal{H}=L^{2}\left(\mathbb{R}^{3}\right) [/itex] and the basis from spherical harmonics.
     
    Last edited: Jun 27, 2005
  5. Jun 28, 2005 #4

    CarlB

    User Avatar
    Science Advisor
    Homework Helper

    Since [tex]L_z[/tex] does not commute with any other [tex]L_u[/tex] for [tex]\hat{u}[/tex] any other direction not parallel or antiparallel to [tex]\hat{z}[/tex], it should be clear that it's pretty easy to arrange for an eigenstate of [tex]L_u[/tex] and [tex]L^2[/tex] to not be an eigenstate of [tex]L_z[/tex].

    In addition, the reverse is also possible. That is, there are eigenstates of [tex]L_z[/tex] that are not eigenstates of [tex]L^2[/tex]. To make one, just take two eigenstates of both [tex]L^2[/tex] and [tex]L_z[/tex] that happen to have the same [tex]L_z[/tex] eigenvalue but different [tex]L^2[/tex] eigenvalues, and mix em up.

    Carl
     
  6. Jun 28, 2005 #5

    dextercioby

    User Avatar
    Science Advisor
    Homework Helper

    I have no idea why i didn't think of linear combinations of eigenstates.:rolleyes:

    Daniel.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Commuting operators
Loading...