1. Limited time only! Sign up for a free 30min personal 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!

Homework Help: Clebsch-Gordan coefficient

  1. Sep 11, 2012 #1
    Hi, for the deutron triplet state, I am trying to find the coefficient using the lowering operator but I kept getting 0 hope someone can help me.

    I can get from |++> to 1/[itex]\sqrt{2}[/itex](|-+> + |+->) but from here I am stuck since j+m=0 for |+->.

    What is wrong with my working? Thanks
  2. jcsd
  3. Sep 11, 2012 #2
    Perhaps you are not remembering that J- has two pieces?

    [tex]J_- = J_{1-} + J_{2-}[/tex]
  4. Sep 12, 2012 #3
    [tex]J_- |+-> + |-+> = J_{1-} + J_{2-}|+-> + |-+>[/tex] I am not sure about this part but i split the operators into
    [tex]J_- = J_{1-}|+-> + |-+> + J_{2-}|+-> + |-+>[/tex]
    so [tex]J_{1-}|+-> + |-+> [/tex] is where I am stuck since [tex]m_1=1/2 , m_2=-1/2 [/tex] so everything will be reduced to 0.
  5. Sep 12, 2012 #4
    Don't forget the brackets!

    [tex]\left( J_{1-} + J_{2-} \right) \left(\lvert +- \rangle + \lvert -+ \rangle \right)[/tex]

    There are 4 terms here, 2 of which survive the lowering operators.
  6. Sep 12, 2012 #5
    Yeah I am aware there are 4 terms but is this correct?
    [tex]J_- = J_{1-}|+-> + |-+> + J_{2-}|+-> + |-+>[/tex]
    If its correct then I should be able to evaluate this [tex]J_{1-}|+-> + |-+> [/tex]
    I am aware that when the lowering operator acts on the lowest eigenvalue it will disappear but I cant continue since the formula has (j+m) term which is 0 for different spins. By the way I am doing revision for this so it would be better if anyone can just tell me what I did wrong here. This is not homework. Thanks
  7. Sep 12, 2012 #6
    You're not using the distributive law correctly: (A+B)(x+y) = Ax + Ay + Bx + By. Every state in the sum should have a lowering operator acting on it.

    Two of the terms vanish but two survive because you're acting on a + with the lowering operator.
  8. Sep 12, 2012 #7
    I really still don't see what I did wrong there. I simply use (A+B)(x+y)=A(x+y)+B(x+y). I am having problem here because I am using the lowering operator on + but I am getting 0 which is clearly wrong but I just don't know why I got it wrong. I will show my working

    Using the lowering operator [tex]J_-|jm>=((j+m)(j-m+1))^{1/2}|jm-1>[/tex]
    on |-+> I got this [tex]J_-|-1/2, 1/2>=((-1/2+1/2)(-1/2-1/2+1))^{1/2}|-1/2,-1/2>[/tex]
    [tex](-1/2+1/2)(-1/2-1/2+1)=(0)(0)=0[/tex] and for |+-> the operator is also acting on the second entry which is - so it goes to 0 as well.
    Clearly I am making a stupid mistake but I really don't see it myself. Hopefully someone can tell me whats wrong
  9. Sep 12, 2012 #8

    [tex]S_{1-} \lvert +- \rangle = \lvert -- \rangle[/tex]
  10. Sep 12, 2012 #9
    Sorry, we cross-posted...

    [tex]\sqrt{(j+m)(j-m+1)} = \sqrt{(1/2 + 1/2)(1/2 - 1/2 + 1)} = 1[/tex]
  11. Sep 12, 2012 #10
    Yes but I believe this is for the case of |++> since j=m=1/2?
  12. Sep 12, 2012 #11
    It's also the case for |+-> because J1- only acts on the +. It leaves the - alone.

    Maybe that's the point you're missing? J1- lowers the first index and leaves the second alone; J2- acts on the second index and leaves the first alone.
  13. Sep 12, 2012 #12
    Yes I am aware of that but in the working I wrote a few post ago, the coefficient is 0 since we have (j+m) which is 0.
  14. Sep 12, 2012 #13
    but j=1/2 and m=1/2.
  15. Sep 12, 2012 #14
    Think this is the part that I got confuse. For |+-> , does it mean that j=1/2 , m=-1/2?
  16. Sep 12, 2012 #15
    Aha! I think we've nailed the problem. What you have here is the tensor product of two spin-1/2 states. So j1 = j2= 1/2. The + and - refer to m1 and m2 respectively.

    For instance |+-> would have j1 = j2= 1/2, m1 = 1/2 and m2 = -1/2.
  17. Sep 12, 2012 #16
    Haha I get it now. Thank you so much for explaining this to me!
  18. Sep 12, 2012 #17
    No problem :cool:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook