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!

Orbital and Spin angular momentum

  1. Apr 5, 2007 #1
    1. The problem statement, all variables and given/known data
    Consider the states for which l=4 and s=1/2.For the state wit hteh largest j, and the largest mj, calculate
    a) the nagle between L and S
    the angle between [itex] \mu_{l}[/itex] and [itex] \mu_{s}[/itex]
    c) the angle between J and the +z axis


    2. Relevant equations
    j=l+/-s

    3. The attempt at a solution
    the largest possible j is 4 +1/2 which is 9/2.
    the largest mj is also 9/2
    since mj ranges from -j to +j.

    now my prof said that this could be done gemetrically that is
    since l=4, then the value of value of L is [itex] \sqrt{4(4+1)}\hbar=3\hbar[/itex]
    the projection on the z axis, that is ml may be -4,-3,...,3,4

    the spin s=1/2, ms=1/2
    the magnitude of the spin angular momentum vector is [itex] \sqrt{3}/2 \hbar[/itex]

    usig the projection of both we can draw the vectors and we can find the nagles between them

    i was jjust wondering if there was a way of doing this using [itex] \vec{L} \cdot \vec{S} [/tex]

    [tex] L \cdot S = |L| |S| \cos \theta [/tex]

    the magnitude of L is calcualted above
    siilarly for S

    but how would one go about calculating L dot S??
    thanks for the help!!
     
    Last edited: Apr 5, 2007
  2. jcsd
  3. Apr 5, 2007 #2

    nrqed

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member


    You can always write
    [tex] \vec{L} \cdot \vec{S} = \frac{1}{2} \bigl( {\vec J}^2 - {\vec L^2} - {\vec S}^2 \bigr) [/tex]
    So this operator sandwiched between states of definite j, l and s gives [itex] \frac{\hbar^2}{2} (j(j+1) - l(l+1) -s(s+1))[/itex]. This is used in the calculation of the spin orbit interaction in hydrogen, for example.
     
  4. Apr 5, 2007 #3

    cgw

    User Avatar

    Someone correct me if I am wrong but:
    The lengths of the vectors (L, S, J) are proportional to sqrt{l(l+1)}, sqrt{s(s+1)}, sqrt{j(j+1)}
    Then maybe the angle between L and J can be had by the cosine law (which to me looks similar to what nrqed has above).
    BTW check your sqrt{4(4+1)}.
     
    Last edited: Apr 5, 2007
  5. Apr 5, 2007 #4
    here we are looking for the largest possible j so

    j = 9/2
    so
    [tex] L \cdot S = \frac{\hbar^2}{2} (j(j+1) - l(l+1) -s(s+1)) = \frac{\hbar^2}{2} \left(\frac{9}{2}\left(\frac{9}{2} +1\right) - 4(4+1) - \frac{1}{2}\left(\frac{1}{2} +1\right) \right) = \hbar^2 [/tex]

    [tex] |L| = \sqrt{20} \hbar [/tex]
    [tex] |S| = \sqrt{3}/2 \hbar [/tex]

    are u sure it should be cosine law or dot product??
     
  6. Apr 6, 2007 #5

    nrqed

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Well, since S, L and J are really quantum operators, one cannot really see them as ordinary vectors. So it does not really make sense to talk about the angle between two operators. But people define the angle between [itex] {\vec L} [/itex] and [itex] {\vec S} [/itex] in a certain state of definite L,S and J as the ratio of the expectation values [tex] \frac{ < {\vec L } \cdot {\vec S} >} { \sqrt{ < {\vec L}^2 >} \sqrt{ < {\vec S}^2>}} [/tex]
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?