# Pentaquark discovery at LHCb

1. Jul 14, 2015

### Staff: Mentor

Pentaquark are particles made out of 4 quarks and 1 antiquark, unlike mesons (quark+antiquark) and baryons (3 quarks). This is by far not the first claim of a pentaquark discovery, but I think it is the most convincing one so far. We'll see which alternative interpretations will come up in the next weeks.
If confirmed, it is a completely new class of hadrons.

Some experimental background:
LHCb looked at the decays of $\Lambda_b^0 \to J/\psi p K^-$. Those four particles are well-known and the mother particle lives long enough to see its decay length: the three daughter particles come from a common point (the decay vertex) that is well separated from the primary interaction point. That allows to reduce the background significantly. LHCb then asked "could this be a decay chain? $\Lambda_b^0 \to X K^-$, $X \to J/\psi p$ with a new particle X"? This would give a peak in the invariant mass distribution of the $J/\psi p$ pair. And indeed they found such a peak.

Not every peak in such a distribution is a particle, which is a serious issue for all discovery claims. There are many effects in a three-body decay that have to be considered because they can lead to bumps in the spectrum.

I see three good arguments why this peak should be a pentaquark (or something even more exotic):
• the peak is quite narrow (first plot in first reference). You can see several other particles contributing to the spectrum in the invariant mass distribution plot, but they are all wide, and even with interference effects it is unlikely that they could give such a narrow peak.
• the Argand diagram (second plot in first reference). It is a plot of the complex amplitude of the unknown contribution as function of the mass. You expect a circle for a particle, and indeed the points are on a circle. There is no reason why it should be circular without a new particle.
• the quark content: the particle is too light to contain a b-quark, but it needs a charm and anticharm quark to produce a $J/\psi$ in the decay. It also needs three (valence) quarks more than antiquarks to be baryon-like. Both together requires at least five quarks.

The arguments are weaker for the second particle which is wider and does not have a clear Argand diagram.

LHCb will certainly try to measure more decay modes, or to find other similar particles.

This discovery is related to the measurement of Z(4430)+, a charged state decaying to $\Psi' \pi^-$.

References:
LHCb press release
arXiv preprint

Abstract of preprint:

2. Jul 14, 2015

### Safinaz

Hi,

I wonder how Standard Model predicteds these five or four quarks particles ? According to the quark model SU(3) has the fundamental representation 3, then we can have octet, decouplet, sextet or 15 representations, while there is no 5 or 4 representations ?

Best.

3. Jul 14, 2015

### resurgance2001

Safinaz
That's a really good question. I would really like to know the answer to this.

From what I have read so far this pentaquark was predicted sometime ago, so I would expect that it is somehow a facet of the standard model. Let's hope some other people chip in with some relevant info.

4. Jul 14, 2015

### Safinaz

I think I knew the answer ..

Like mesons which are $q \bar{q} ~$ SU(3) singlet states, arising from the tensor product $3 \times 3^* = 1+8$ ,

One of the new pentaquark particles $^*$ consisting of two up quarks, a down quark, a charm quark and an anti-charm quark, so now we will have $3\times 3 \times 3 \times 3 \times 3^*$ , which i think it can give SU(3) singlet, is it correct ? that's also as baryons states:
$$3 \times 3 \times 3 = 1 + 8 + 8 + 10$$

The four-quark exotic particles like $Z_c(3900)$ consisting of up, anti-down, charm and anti-charm quarks , so we have $q \bar{q} q \bar{q}$ state or $3 \times 3^* \times 3 \times 3^*$ .

The question now how many exotic particles have not been found yet and are allowed or predicted by the quark model ?

* I think it's called Pc(4450)+, look for example :
http://www.symmetrymagazine.org/article/july-2015/lhc-physicists-discover-five-quark-particle

Best.

5. Jul 14, 2015

### Lord Crc

As a layman the LHCb appears to be "the little detector that could". Ok it's not that small but still, it's seemingly been churning out interesting results for a while and presumably with more to come in run 2.

6. Jul 14, 2015

Staff Emeritus
I happen to be in a meeting with a number of colleagues I don't usually talk to. My reaction was "Maybe this time it's real".

I am not convinced by the peak. I've seen better looking peaks go away.
I find the Argand plot strong evidence. It's hard - not impossible, but hard - to generate a spurious pole.
I find the Dalitz plot good evidence. It does not look like phase space sculpting.

The second and third points reinforce each other. I don't find the second peak nearly as compelling.

People are asking whether this is a "real pentaquark" or only a p-Psi bound system. I am not sure this question even has any meaning.

7. Jul 15, 2015

### atyy

What is phase space sculpting?

8. Jul 15, 2015

Staff Emeritus
The acceptance of the detector can be reflected into kinematic quantities, like mass. In extreme cases it can enhance or even manufacture a peak in a mass distribution.

9. Jul 15, 2015

### Staff: Mentor

If there are exited states, one description could be more meaningful than the other. I guess the excited states of hadrons look different from the spectrum you would expect with pions as exchange particles.
Not sure if those states (if they exist at all) would clearly favor one description, however.

10. Jul 15, 2015

### BiGyElLoWhAt

Just jumped on to link the paper. I haven't read the paper yet, but I'm going to assume that about half of the jargon will go right over my head. A quck question for someone that knows more about this area:

How can you distinguish between a gluon absorbed by a quark, the quark being ejected, and a quark anti quark pair being generated, resulting in a baryon meson pair (which is allegedly what this penta quark decays into)? With the time frame that this particle lives for, I don't see how you could distinguish between the 2 situations (pentaquark vs. J/phi meson + proton baryon)

11. Jul 15, 2015

### Staff: Mentor

The invariant mass distribution of the J/Psi + proton shows a structure, and that structure looks very much like a bound QCD state - a particle (well, two of them).

An interesting point: LHCb didn't look specifically for this decay, which means their triggers were probably not optimized for it. Chances are good they threw away most of those events - the data rate is way too massive to keep everything. They will certainly optimize their trigger to study those new particles now, so the dataset could get larger quickly.

Edit: No, they just used the displaced J/Psi trigger. Hard to improve that...

Last edited: Jul 15, 2015
12. Jul 15, 2015

### BiGyElLoWhAt

I guess the question lies in your first paragraph. How do you tell the difference between a bound quantum chromdynamics state consisting of 5 quarks versus 2 bound states consisting of 2 quarks and 3 quarks? This is the part, to me, that seems it would be veerrry hard to distinguish experimentally.

13. Jul 15, 2015

### Safinaz

Is there a limit for how many exotic particles have not been found yet and are allowed or predicted by the quark model ?

14. Jul 16, 2015

### Lord Crc

Potentially silly question: Will any of the extensions to the SM (supersymmetry etc) be constrained in any significant way by this result? Or a refinement? Assuming they did indeed find pentaquarks.

15. Jul 16, 2015

Staff Emeritus
No.

16. Jul 16, 2015

### Lord Crc

Thanks, figured it would be a long shot but had to ask.

17. Jul 20, 2015

### websterling

In the decay of the $\Lambda_b^0$ to a kaon and a pentaquark, what is the origin of the $u \bar{u} ~$ pair without a connected vertex? Is it the decay of a gluon whose line is not shown in the drawing? Or is it something else?

Thanks

18. Jul 20, 2015

### ChrisVer

It can be more than a gluon I guess (a photon could be)... but gluons are the dominant ones

19. Jul 20, 2015

### PineApple2

Does this discovery reveal new physics, besides the assembly of new particles made of more quarks? i.e. does it change or give new information about current physical models or theories?

20. Jul 20, 2015

### BiGyElLoWhAt

I think with a little more study it could help us figure out what quark combinations are viable. We've seen 2, 3, and now 5 quark combos. Is a 7 quark (5q 2 anti) viable? It could still produce a color neutral object

21. Jul 20, 2015

### ohwilleke

Yes, although it isn't a finite limit, just one that distinguishes between allowed and not allowed possibilities.

The possibilities start with a basic rule. In general, QCD permits color charge neutral combinations of QCD color charged entities, but not other combinations of QCD color charged entities. This rule, in principle, allows for (1) glueballs (with no quarks), (2) quark-antiquark mesons, (3) and (4) three quark or three anti-quark baryons, (5) two quark and two anti-quark tetra quarks, (6) and (7) four quark and one antiquark or four antiquark and one quark pentaquarks, etc. ad infinitum. So long as a particle has 3N net quarks (with "net quarks" equal to quarks minus antiquarks), for N=any integer, the color charge neutrality rule is satisfied.

The question is whether this naive application of just one of the most basis top level rules of QCD is sufficient to establish that QCD really permits combinations other than mesons or baryons. It could be that a "true tetraquark" or "true pentaquark" is forbidden or profoundly suppressed relative to meson molecule or meson-baryon molecule alternatives by some other non-obvious aspect of QCD.

A "true tetraquark" or a "true pentaquark" or a "true hexaquark" implies that all of confined within a single confinement space, while meson molecules, meson-baryon molecules, or baryon-baryon molecules (common practice is to call these molecules, but unstable atomic nuclei held together mostly by pion exchange are a more accurate analogy) have two confinement spaces and a bond between them.

One way to distinguish between a true pentaquark and a mere meson-baryon molecule that has been proposed is that the width of the resonance of the composite object in the two respective possibilities. A true pentaquark and a hadron molecule with the same quark content ought to have a different resonance width.

Another is to do the very hard math to calculate the expected mass of the composite object relative to the hadron molecule alternative. Naively, one would expect that a different amount of binding energy is necessary to hold four or five quarks together in a single confinement space, than is necessary to bind the quarks of two hadrons into two separate confinement spaces and then to add the binding energy of the mostly pion mediated bond between the two hadrons. If the binding energy of a true pentaquark was much greater than the combined binding energy in a meson-baryon molecule, then it ought to be possible to calculate a suppression factor of one possibility v. the other and to measure the difference in individual resonances. But, given tunneling possibilities and the possibility that extra binding energy in a true pentaquark just gets converted to kinetic energy when it all falls apart, this might be pretty hard to distinguish observationally, relative to differences in resonance width. You'd really have to look a chart of resonances v. momentum exchange magnitude (Q^2) to distinguish the two possibilities statistically from a large data set of such resonances, and you'd need good theoretical calculations to distinguish between what was expected in each scenario.

From the perspective of "New Physics" (or at least "new insights into Standard Model QCD"), the boring possibility (that we are seeing hadron molecules rather than true tetraquarks or true pentaquarks) is actually the more interesting one. If true tetraquarks and true pentaquarks don't actually happen, then one needs either New Physics, or less remarkably, new insights into Standard Model QCD, to explain why composite quark combinations that are naively possible do not exist in reality, or alternatively, are dramatically suppressed relative to hadron molecule states.

Why, for example, don't we routinely see true hexaquark states in large atomic nuclei (which our experience teaches us is made up almost entirely of boring old protons and neutrons that are almost never unfaithful to their confined threesomes)? Nothing obvious about QCD at a qualitative level explains why this should be so. True tetraquarks and true pentaquarks aren't really "new physics" at all, even though they are exotic and haven't been seen before (unless we've finally seen one now). We've been expecting them to be out there for even longer than we've been looking for the Higgs boson.

But, for example, if one learned that the binding energy of a true pentaquark was on the order of 100 times that of a meson-baryon molecule, this would explain why pentaquarks, while possible in principle, are so rarely seen. Then, we might need, for example, a machine 10-100 times as powerful as the LHC to see a true pentaquark, which might be only 1 in 10,000 times as common as hadron molecules with the same quark content.

In the case of "New Physics" there might simply be a new rule of QCD, call it the "strong confinement rule" that states that not only are all quarks and gluons confined, but that all quarks and gluons are confined in either quark-antiquark mesons, or in three quark baryons, or in three antiquark baryons (a rule that would also forbid glueballs which have been exceptionally elusive to date as well despite having properties calculated quite exactly since the early 1980s with predicted masses that should be easily accessible).

22. Jul 21, 2015

### rollingstein

Naive question: Just like when data mining Big Date you are liable to detect artifacts once in a while that might look like trends, when the LHC sees a circular Argand etc. how many potential other non-special events had it detected?

Like isn't there the possibility that after a huge number of attempts just a pure lining up of coincidences gets you something that looks like a clue?

Or does the statistics take care of compensating for this?

23. Jul 21, 2015

Staff Emeritus
Classical statistics does a pretty good job of answering the question "what is the significance of this effect". On top of that, one needs to consider the "trials factor", sometimes called the "look-elsewhere effect". That answers the question you asked - what is the effect of doing many simultaneous tests. The exact value for this is debated every time it comes up, and is part of the art of statistics rather than the science, but it is usually measured in the dozens, not thousands.

24. Jul 21, 2015

### rollingstein

Can you elaborate some more on this? What's the difference between the two effects you are describing? Isn't the second part what is corrected by a Bonferroni multiple comparisons correction or similar?

Also, in this specific pentaquark case, what are the order of magnitude numbers: i.e. How many events did they look at before they found an approximately circular Argand? And what's the probability of seeing what they say by statistical, probabilistic co-incidence?

25. Jul 21, 2015

### Staff: Mentor

It will help to understand QCD (quantum chromodynamics) better, which is also useful for calculations related to mesons and baryons.

Concerning the look-elsewhere effect: LHCb published about 270 papers so far. Not all of them are physics results, some of them have multiple measurements combined, so several hundred physics results is probably a reasonable estimate. You would expect one or two three-sigma effects somewhere just by random chance. Actually, the $\Delta A_{CP}$ charm mixing measurement was 3.5 sigma and went away with more data. The P5' thing is a ~3.7 sigma deviation from predictions, depending on what exactly you look at.
Yes, the large number of studies leads to some fluctuations. This is known, and taken into account when interpreting the results. Despite the large number of studies, the random chance to see a 5 sigma effect is very small already. The random chance to see 9 sigma is negligible.