Has photoproduction of gluons been observed?

[ohwilleke]

TL;DR

I am wondering if a process with an initial state of two photons leading to an end state with two (or more) gluons has been observed.

I am wondering if a process with an initial state of two photons leading to an end state with two (or more) gluons has been observed.

I have seen some papers (like this one) that seem to suggest that this is the case, but none of them say so as clearly as I would like to be confident that this is true.

 

[Vanadium 50]

Since nobody has seen a free gluon, the answer to your question as posed is no.
If you had a different question in mind, you should ask it.

 

[MathematicalPhysicist]

Since nobody has seen a free gluon, the answer to your question as posed is no.
If you had a different question in mind, you should ask it.

Well, no one can "see" any particles, we detect resonance peaks which we interpret as a particle.

 

[Vanadium 50]

You're not helping. We can see protons, for example, just fine.

 

[MathematicalPhysicist]

You're not helping. We can see protons, for example, just fine.

Do you want to say you see them directly?
You see an image, but it doesn't mean you can see them without your equipment.
Anyway, I searched google for "image of a proton" and couldn't find one, do you have one?

 

[ohwilleke]

Since nobody has seen a free gluon, the answer to your question as posed is no.
If you had a different question in mind, you should ask it.

I think this proves too much. By that reasoning, no one has ever observed an up, down, strange, charm, or bottom quark, since there is likewise no evidence of free quarks that are not top quarks (in each case using the term quark to include antiquarks in this sentence).

But, of course, in a reasonable interpretation of what "observed" means in this context, we have seen, for example, experimental evidence of photoproduction of some quark-antiquark pairs (e.g. here), a process which is simply the inverse of quark-antiquark annihilation, which is one of the first things taught in quantum mechanics.

There are all sorts of particles in high energy physics that one only observes indirectly, for example, through their decay products, after statistically removing known background "noise" from other know processes, for example. But we still "observe" those particles.

"Observed" does not literally mean "see" with your technologically enhanced eyes, in this context. Instead, it means that there is strong experimental evidence that it has occurred (five sigma is a bit arbitrary, but the usual standard for such things, however, if there were 3 or 4 sigma experimental evidence for the existence of such a process that would be worth mentioning too). Observe is in contrast to merely theoretically predicting that something exists without experimentally evidence supporting its existence, like the Higgs boson at 125 GeV pre-2012.

The question is whether there is strong experimental evidence supporting the existence of the process or not.

 

[mfb]

I'm with V50 here. The experimental measurement will always be hadrons. Which hadrons are you looking for? We have seen photoproduction of hadrons, and diagrams with gluons contribute to them.

 

[Vanadium 50]

Do you want to say you see them directly?

You're really not helping. The fact that protons are really, really tiny has nothing whatsoever to do with the question asked.

 

[Vanadium 50]

I show you an event $\gamma \gamma \rightarrow \pi^+ \pi^-$ and you say "Nope. Coulda been quarks." Then I show you an event $\gamma \gamma \rightarrow \rho^+ \rho^-$ and you say "Nope. Coulda been quarks." This game will get very old very quickly.

 

[ohwilleke]

I show you an event $\gamma \gamma \rightarrow \pi^+ \pi^-$ and you say "Nope. Coulda been quarks." Then I show you an event $\gamma \gamma \rightarrow \rho^+ \rho^-$a nd you say "Nope. Coulda been quarks." This game will get very old very quickly.

What I am imagining (the numbers in the example are just made up) is something more like the SM predicts 6 events in a very large data set from quark production from $\gamma \gamma \rightarrow \pi^+ \pi^-$ and 3 events from quark production from $\gamma \gamma \rightarrow \rho^+ \rho^-$ and 2 events from other cases of photoproduction of quarks, for a total of 11 hadronic events predicted with an uncertainty of ± 1.

If the process $\gamma \gamma \rightarrow gg \rightarrow \pi^+ \pi^-$ and from other $\gamma \gamma \rightarrow gg \rightarrow hadron^+ hadron^-$ decay channels existed, however, the SM prediction would be an additional 7 hadronic events predicted with an uncertainty of ± 0.5, in particular proportions. Obviously a real experiment would involve a lot more technical detail, but that would be the basic kind of experiment I'm thinking about conceptually.

We look at our data set and identify 17 events with a distribution by hadron type which is close to the predicted mix of hadrons predicted for the combined processes in a Chi-square test.

This would be strong observational evidence of photoproduction of gluons.

What I am curious about is whether an experiment has been done and produced results like that which provide that kind of strong evidence.

Also, for that matter, I'm not entirely clear about whether the SM predicts a detectible number of events from such processes at all. In the alternative, I'd also be interested to know if someone calculated SM predictions and the predicted number of events in a scenario like the one above would be so small that the kind of experiment I use as an example wouldn't come close to being able to see particles created through the channel with current or near future technology, since if true, that would explain why no one has bothered to try to do that experiment.

 

[Vanadium 50]

OK, that's at least answerable. The answer is "no".

1. In quantum mechanics, we add amplitudes and then square. Furthermore, the amplitudes may have different phases. Adding a new amplitude doesn't even necessarily make the probability go up.
2. The process γγ→hadrons occurs at order α² and the process γγ→gg → hadrons occurs at order α²α_s². That's 50-100x smaller. We cannot predict γγ→hadrons to that degree of accuracy.
3. The dominant effect is not γγ→gg → hadrons but the interference between the non-gluonic and gluonic pieces. (a consequence of adding amplitudes) A discrepancy between the measurement and the pure QED prediction does not require adding the process of interest.
4. Furthermore, there is a non-perturbative piece to the calculation called the form factor, which is essentially the wavefunction overlap of the constituent partons, and this is determined from experiment, A discrepancy would then be absorbed into the form factors.

 

[ohwilleke]

OK, that's at least answerable. The answer is "no".

1. In quantum mechanics, we add amplitudes and then square. Furthermore, the amplitudes may have different phases. Adding a new amplitude doesn't even necessarily make the probability go up.
2. The process γγ→hadrons occurs at order α² and the process γγ→gg → hadrons occurs at order α²α[_sSUP]2[/SUP]. That's 50-100x smaller. We cannot predict γγ→hadrons to that degree of accuracy.
3. The dominant effect is not γγ→gg → hadrons but the interference between the non-gluonic and gluonic pieces. (a consequence of adding amplitudes) A discrepancy between the measurement and the pure QED prediction does not require adding the process of interest.
4. Furthermore, there is a non-perturbative piece to the calculation called the form factor, which is essentially the wavefunction overlap of the constituent partons, and this is determined from experiment, A discrepancy would then be absorbed into the form factors.

Thanks. That's helpful and answers the question I had.

 

[arabusov]

OK, that's at least answerable. The answer is "no".

1. In quantum mechanics, we add amplitudes and then square. Furthermore, the amplitudes may have different phases. Adding a new amplitude doesn't even necessarily make the probability go up.
2. The process γγ→hadrons occurs at order α² and the process γγ→gg → hadrons occurs at order α²α_s². That's 50-100x smaller. We cannot predict γγ→hadrons to that degree of accuracy.
3. The dominant effect is not γγ→gg → hadrons but the interference between the non-gluonic and gluonic pieces. (a consequence of adding amplitudes) A discrepancy between the measurement and the pure QED prediction does not require adding the process of interest.
4. Furthermore, there is a non-perturbative piece to the calculation called the form factor, which is essentially the wavefunction overlap of the constituent partons, and this is determined from experiment, A discrepancy would then be absorbed into the form factors.

Nevertheless, γγ→gg crosssection has been already studied theoretically (for example here: https://arxiv.org/abs/1310.6701), and at some point I guess the effect won't be negligible at least in ultraperipheral collisions at LHC.