PDA

View Full Version : Six Generators of Group SO(4)


YOBDC
Dec15-10, 12:14 AM
Recently,I have read a article which refered "the six generators of group SO(4)".And who can tell me what are these generators and what are their matrix forms?

tom.stoer
Dec15-10, 12:23 AM
There a three rotations La (generating the SO(3) subgroup of SO(3,1) and three boosts Ka.

In addition one can complexify the algebra to SO(4,C) and define the new linear combinations

L±a = La ± iKa.

These new sets of generators define two mutually commuting SU(2,C) algebras.

The matrix form can e.g. be found in "Jackson: Electrodynamics".

DrDu
Dec15-10, 01:59 AM
The Hamiltonian of the Coulomb problem is invariant under SO(4) and is discussed e.g. in Landau Lifshetz, Quantum Mechanics. In that case the six generators are the three generators of rotations and the three so called Runge Lenz vectors which describe the orientation of the big axis of the ellipses of the particles.

arkajad
Dec15-10, 03:24 AM
They are:
L_{12}=\begin{pmatrix}0&1&0&0\\-1&0&0&0\\0&0&0&0\\0&0&0&0\end{pmatrix}

and five similar L_{ij},\quad i<j

YOBDC
Dec15-10, 03:36 AM
They are:
L_{12}=\begin{pmatrix}0&1&0&0\\-1&0&0&0\\0&0&0&0\\0&0&0&0\end{pmatrix}

and five similar L_{ij},\quad i<j

Did you mean that for each Lij(i<j),there are only two non-zero elements:lij=1 and lji=-1?

arkajad
Dec15-10, 04:00 AM
Did you mean that for each Lij(i<j),there are only two non-zero elements:lij=1 and lji=-1?

Yes. But, of course, one can always choose a different basis of generators by taking independent linear combinations of the above ones,

YOBDC
Dec15-10, 04:22 AM
Yes. But, of course, one can always choose a different basis of generators by taking independent linear combinations of the above ones,

Oh, I see. These generators must be anti-symmetric. Thank you very much!

tom.stoer
Dec15-10, 04:57 AM
Note that the counting is different; arkajad uses ij with i<j (ij being spacetime indices) whereas I am using a=1..3 (NOT being spacetime indices; suppressed). Of course in both cases you have six generators in total and of course they are one-to-one.

samalkhaiat
Dec15-10, 04:54 PM
These generators must be anti-symmetric.

You can represent the n(n-1)/2 generators of any SO(n) in some orthonormal base, by the operator

L_{nm} = |n\rangle \langle m| \ - |m \rangle \langle n|

where n,m = 1, 2, ...,n and;

\langle n|m\rangle = \delta_{nm}

From that you can find the location of the non-zero matrix elements by calculating the martix element;

\langle r|L_{nm}|s \rangle