On srednicki page 240 (print) there is a derivation of the Gordon identity, and it starts with stating that
\require{cancel} \gamma^{\mu}\cancel{p} = \frac{1}{2} \big\{\gamma^{\mu},\cancel{p} \big\} + \frac{1}{2} \big[ \gamma^{\mu},\cancel{p} \big] = -p^{\mu} -2iS^{\mu\nu}p'_\nu{}
Where exactly does this come from?
Answer
Let \gamma^\mu be a gamma matrix, i.e., \frac{1}{2}(\gamma^\mu\gamma^\nu+\gamma^\nu\gamma^\mu)=-\eta^{\mu\nu} \tag{1}
Let S^{\mu\nu}\equiv\frac{i}{4}(\gamma^\mu\gamma^\nu-\gamma^\nu\gamma^\mu) \tag{2}
Let us write \gamma^\mu\gamma^\nu=\frac{1}{2}(\gamma^\mu\gamma^\nu+\gamma^\mu\gamma^\nu) \tag{3} (note that I did nothing yet)
Next, write 0=\frac{1}{2}(\gamma^\nu\gamma^\mu-\gamma^\nu\gamma^\mu) \tag{4} and add it to (3): \gamma^\mu\gamma^\nu=\frac{1}{2}(\gamma^\mu\gamma^\nu+\gamma^\mu\gamma^\nu)+\frac{1}{2}(\gamma^\nu\gamma^\mu-\gamma^\nu\gamma^\mu)=\frac{1}{2}(\gamma^\mu\gamma^\nu+\gamma^\nu\gamma^\mu)+\frac{1}{2}(\gamma^\mu\gamma^\nu-\gamma^\nu\gamma^\mu) \tag{5} where I just rearranged the terms.
Finally, plug (1) and (2) into (5): \gamma^\mu\gamma^\nu=-\eta^{\mu\nu}-2iS^{\mu\nu} \tag{6}
This is all we need to prove the relation given by Srednicki: \gamma^\mu\not p\equiv p_\nu\gamma^\mu\gamma^\nu=p_\nu(-\eta^{\mu\nu}-2iS^{\mu\nu})=-p^\mu-2iS^{\mu\nu}p_\nu \tag{7}
Note that this can be slightly generalised: let a^\mu,b^\mu be any two objects (vector, matrices, operators, etc.)
With this a^\mu b^\nu=\frac{1}{2}\{a^\mu,b^\nu\}+\frac{1}{2}[a^\mu,b^\nu] \tag{8} as can be easiliy checked by expanding the commutators. In this case, a^\mu=b^\mu=\gamma^\mu, so that (6) follows after realising that \{\gamma^\mu,\gamma^\nu\}=-2\eta^{\mu\nu} by definition of the gamma matrices, and [\gamma^\mu,\gamma^\nu]=-4iS^{\mu\nu} by definition of S^{\mu\nu}.
No comments:
Post a Comment