Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Tricky Pauli matrices

  1. May 27, 2009 #1
    I am trying to recover the following calculation (where [tex]K,A[/tex] are 4x4 matrices in SL(2,C)):

    "We expand [tex]K'=AKA^{\dagger}[/tex] in terms of [tex]k^a[/tex] and [tex]k'^{a}=(\delta_a^{b} + \lambda_a^{b} d\tau)k^b[/tex]. Multiplying by a general Pauli matrix and using the relation [tex]\frac{1}{2}tr(\sigma_{a}\sigma_{b})=\delta_{ab}[/tex] yields the expression:
    \lambda_b^{a} = \frac{1}{2}\eta^{ac}tr(\sigma_{b}\sigma_{c}A+\sigma_{c}\sigma_{b}A^{\dagger})

    I have been playing with the relations for a while but I guess I miss some knowledge on the properties of Pauli matrices because I don't manage to find the result. In particular, what would the "expansion" of [tex]AKA^{\dagger}[/tex] (which I guess is necessary here?) look like in Einstein summation notation ? Any help would be extremely appreciated!
  2. jcsd
  3. May 30, 2009 #2
    the answer lies on wikipedia.
  4. May 30, 2009 #3


    User Avatar
    Science Advisor
    Homework Helper

    Well, give a link, then :rolleyes:which article? :smile:
  5. May 30, 2009 #4
    Well I had of course already checked beforehand the wikipedia page on Pauli matrices (http://en.wikipedia.org/wiki/Pauli_matrices) but had not found a relation to solve this problem... So Ryuunoseika, which article are you talking about ?
  6. May 30, 2009 #5

    George Jones

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Something here seems either not quite right or incomplete.

    If [itex]k'^{a} = \left(\delta_b^{a} + \lambda_b^{a} d\tau \right) k^b[/itex] (careful with index placement) and [itex]d \tau[/itex] is infinitesimal (?), then [itex]k'[/itex] and [itex]k[/itex] differ by an infinitesimal amount, so the transformation is an infinitesimal version of [itex]K'=AKA^{\dagger}[/itex]. Then, the sum in final result could come from the product rule.

    I'm just guessing. More context is needed.
  7. May 30, 2009 #6
    Dear George,

    Thanks a lot for your answer. First, yes sorry I misplaced the indices in the first relation, the correct relation is:
    [tex]k'_{a}=(\delta_a^{b} + \lambda_a^{b} d\tau)k_b[/tex]

    Concerning [tex]d\tau[/tex], it is actually a [tex]\delta u[/tex], where [tex]u[/tex] is the affine parameter along the trajectory of a photon.

    But what do you mean by "product rule" ? Do I have to develop explicitely [tex]AKA^{\dagger}[/tex] in indices notation and try to recover at the end the [tex]A[/tex] and [tex]A^{\dagger}[/tex] which appear in the trace ?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook