A proton-proton scattering event cannot take place with the exchange of a single gluon. The argument is that an attempt to draw a Feynman diagram for this process results in colourful outgoing states with the exchange of a single octet. I'm just wondering,

  1. What is then wrong with the following diagram?

enter image description here

I realise $b \bar b$ does not exist alone as one of the eight gluons (there is no linear combination of the eight orthogonal states to get a pure $b \bar b$ gluon state) but I could write the exchanged gluon as, say, $b \bar b - g \bar g$ which is a viable gluonic state and as far as I can see would not violate colour conservation in the above diagram.

2) I've seen in many sources that the colour factor $C_F$ for the amplitude between two colour singlet states mediated between a single gluon is given to be $4/3$. It is then said this results in an interaction potential between two singlets to be $-4/3 r^{-1}$. But if this process does not exist (as demonstrated e.g in the context of the proton proton scattering) then what is the meaning of this non vanishing colour factor - should it not be identically zero since the process is not feasible?

See such remarks made in e.g, at the beginning of this video, https://www.youtube.com/watch?v=_VB-pYuIZw4 and at 55:30.


For your first question, the top group does have color after the exchange. It began with no color and ends up with color 'blue anti-blue'. That state doesn't have any net color charge, but it still has color.

As an example, suppose we have a group of particles with spin 0 and we add another particle with spin $1/2$. If the added particle is in the state $|\uparrow\rangle + |\downarrow\rangle$, then the new state has no net spin in the $z$ direction. But that doesn't mean it's spinless -- it has a spin of $1/2$. You can tell the difference by how the state transforms under rotations; similarly the blue-antiblue state can be distinguished from the true no color state by how it transforms under color rotations.

For your second question, this seems to be a misunderstanding. The example is about the interaction between two quarks which each have color, but whose net color is a singlet. It's not about the interaction of two distinct singlets.

  • $\begingroup$ thanks for your answer! I understand that bbbar is not a colour singlet but I don’t see why the top group has a net colour? The left and right hand side have quarks with r,g,b content? Since a gluon is a colour octet, it makes sense that the left and right sides are not both colourless but how to see this on the level of Feynman diagram colour labels? $\endgroup$ – CAF Jan 12 '18 at 23:55
  • $\begingroup$ @CAF The issue is that there’s no such thing as a pure blue anti blue gluon. Such a gluon corresponds to a matrix with trace 1, but the gluon matrices all have to have zero trace; one basis for them is the Gell-Mann matrices. $\endgroup$ – knzhou Jan 13 '18 at 7:37
  • $\begingroup$ So e.g one permissible gluon is a rrbar-bbar exchange. There is no fermion flow with r to compensate the rrbar so this amounts to a colourful exchange. That correct? $\endgroup$ – CAF Jan 13 '18 at 10:53
  • $\begingroup$ @CAF That's right, you end up with net color "r rbar" which is in the octet, not in the singlet. $\endgroup$ – knzhou Jan 13 '18 at 10:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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