I'm reviewing page 59 of the QFT notes here and am a little confused by a reference frame argument. You can compute the second order probability amplitude term for nucleon-nucleon scattering to be


in a scalar field theory approximation. Now the author argues that we may remove the $i\epsilon$ terms by moving to the centre of mass frame.

Here he says that $p_1=-p_2$ in this frame and that $|\vec{p_1}|=|\vec{p_1'}|$ by conservation of momentum. He continues to claim that the four-momentum of the meson is hence $k=(0,\vec{p}-\vec{p'})$ so $k^2<0$. Quite what $p,p'$ are I don't know exactly.

I don't understand this argument at all. Surely in the centre of mass frame, the sum of all momenta $p_i$ and $p_i'$ is zero (.)? Also where has the second constraint come from? I don't see how morally you could get more than my claim (.).

Could someone explain this argument to me? Very many thanks in advance.

  • $\begingroup$ Be careful about distinguishing three- and four-vectors. Using boldface letters for 3-vectors, we have in the CM frame $\mathbf{p_1} + \mathbf{p_2} = 0$ (as 3-vectors), but $p_1^0, p_2^0 > 0$ (you can compute these using the on-shell condition), so $p_1 \neq -p_2.$ If you write $p := |\mathbf{p_1}| = |\mathbf{p_2}|,$ $p' := |\mathbf{p'_1}| = |\mathbf{p'_2}|$, then you can show that $p = p'$ due to conservation of energy (please do this carefully). $\endgroup$
    – Vibert
    Commented Dec 27, 2012 at 14:02

1 Answer 1


The transferred 4-momentum $k = p - p'$ is a difference (not a sum) and is a momentum of the meson. In the CM reference frame it has only space coordinates, so its square is negative: $k^2=0^2-(\vec{p}-\vec{p}')^2=-(2\vec{p})^2<0$.

  • $\begingroup$ I think my problem is that I don't know how $p$ is related to $p_i$ and $p_i'$. Could you expand on that? And why does it have only space coordinates in the centre of mass frame? I understand if we're talking about the centre of mass for the meson itself, but aren't we interested in the centre of mass for the whole system? Sorry if I'm missing something simple! Many thanks. $\endgroup$ Commented Dec 27, 2012 at 13:56
  • $\begingroup$ The system is a couple of nucleons with the same mass. 4-momentum conservation law reads: $p_1+p_2=p'_1+p'_2$. The transfered momentum is by definition $k=p_1-p'_1$. In the CMRF $\vec{p}_1=-\vec{p}_2$. $\endgroup$ Commented Dec 27, 2012 at 14:03
  • $\begingroup$ So in his exposition and your answer there is a subscript 1 missing from the $p$ then? Or have I missed something. Thanks a lot. $\endgroup$ Commented Dec 27, 2012 at 14:06
  • $\begingroup$ Yes, you can add a subscript 1 to $p$. Omitting the subscript is due to fact that nearly the same expression occurs to $p_2$ too because of 4-momentum conservation. $\endgroup$ Commented Dec 27, 2012 at 14:10
  • $\begingroup$ Ah right that makes much more sense now. I still don't see why the first components of $p_1$ and $p_1'$ are necessarily the same though. I presume it's energy conservation, but why are you allowed to just do that locally? $\endgroup$ Commented Dec 27, 2012 at 14:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.